The US automaker had expanded the use of AI across its operations, including quality inspections, aiming to improve productivity and reduce costs.

However, executives said automated systems did not deliver the expected results.

“Artificial intelligence is a fantastic tool, but it’s only as good as the information you use to train it,” Charles Poon, Ford’s vice president of vehicle hardware engineering, told reporters.

Poon said the company had underestimated the value of experienced engineers, many of whom had worked through multiple vehicle development cycles.

He said Ford has since brought back more than 300 veteran quality inspectors to help train AI systems and mentor younger employees.

“We recognised that for us to enhance some of our automation and machine learning and artificial intelligence tools, we needed to ensure they were trained by the most experienced individuals,” he said.

Ford had deployed AI-powered cameras at its manufacturing plants to identify quality issues and reduce supply chain disruptions.

However, Poon acknowledged that relying solely on AI and design data did not produce the desired quality standards.

The company’s comments came as Ford regained the top spot among mainstream automakers in the US J.D. Power Initial Quality Study for the first time since 2010.

Ford said its quality improvements were supported by a broader talent overhaul, including leadership changes across engineering, manufacturing and supply chain operations, along with the return of experienced engineers.

The exposure threatens the carefully negotiated business of building the iPhone, which Apple assembles from a thicket of suppliers worldwide.

It could also upset Apple and its relationship with Tata, given most of the supplier arrangements are fiercely protected by Apple, and could also hand ​rivals, counterfeiters and its own vendors a view of who makes what.

Tata, which both supplies parts and assembles iPhones as a contract manufacturer, is emerging as ​one of Apple’s most important manufacturing partners outside China, an expansion that is a cornerstone of Prime Minister Narendra Modi’s push to ⁠make India an electronics manufacturing powerhouse.

Apple is reportedly on track to release its iPhone 18 Pro and Pro Max in September.

The leak comes at a difficult time ​for Apple, which last week raised iPad and MacBook prices due to soaring memory and storage chip costs, with analysts expecting Apple to increase iPhone prices in the coming months.

Reuters ​has previously reported the Tata Electronics leak of more than 200,000 files on the dark web by World Leaks, which had files with purported component design papers of older iPhones and some parts of Tesla - both Tata clients.

They also included documents of Taiwan Semiconductor Manufacturing Co and Qualcomm, both of which make parts used in iPhones.

New documents reviewed by Reuters show there are at least six files that map ​many components in the iPhone 18 Pro models to the specific company that supplies them.

These include details of chips on its main circuit board and parts of ​the battery and cameras.

Apple considers this detail sensitive and is concerned about the documents being shared on the dark web as they relate to unreleased models, according to the person familiar with ‌the matter.

⁠The data maps suppliers to iPhone parts, which Apple does not disclose in its public database of suppliers, the person added.

In all, the documents detail hundreds of parts to be on the upcoming iPhone 18 Pro models.

The records also show where Apple draws a part from several suppliers and where it relies on just a few, laying bare both its bargaining leverage and its vulnerabilities.

Spokespeople for Apple and Tata did not respond to Reuters queries.

World Leaks has previously claimed responsibility for a Nike break-in.

Reuters has not verified the ​authenticity of the data and could not ​immediately reach World Leaks for comment.

News website ⁠AppleInsider first reported last week that iPhone 18 Pro documents were part of the Tata leak.

Reuters has previously reported that Apple is investigating the matter and working with Tata on long-term measures.

Tata has restricted internal access to sensitive systems as it investigates the ​leak, and hired a global consultant to conduct a forensic audit.

Drop-test images

Several of the leaked files carried Apple “confidential” watermarks and internal ​Apple code-names consistent with ⁠the iPhone 18 Pro generation, according to the source familiar with the matter.

Inside the folder for iPhone 18 Pro files are photographs of iPhones undergoing drop tests at one of Tata’s plants, dated early 2026.

They depicted a conventional slab-shaped, grey handset with a three-rear-camera setup and the Apple logo.

Reuters could not with certainty identify the model number of the phone, ⁠but the source ​said the photos are of iPhone 18 Pro models.

For Apple and Tata, the breach cuts at the ​trust underpinning their partnership. Apple’s move into India rests on its newest major assembler, Tata, just as the company increasingly diversifies beyond China.

The bet has fast paid off: India is on track to make 26% of ​the world’s iPhones in 2026, up from 6% four years ago, according to Counterpoint, a research firm.

The announcement would mark President Lee Jae Myung’s boldest push yet to align South Korea’s AI and chip ambitions with his pledge to narrow regional disparities and revive economies beyond the Seoul metropolitan area.

Lee will preside over the event, framed ​as a national “great leap” due to be unveiled around his office, said, with ministries covering industry, science, climate and ​transport set to outline policy support.

Samsung Electronics and SK are expected to present investment plans, and their chairmen, ⁠Jay Y.

Lee and Chey Tae-won are among the business leaders tipped to attend by local media.

Representatives of other firms, including LG Electronics, HD ​Hyundai Robotics, Korea Electric Power Corp and Korea Water Resources Corp are also attending, Lee’s office said.

The package will span semiconductors, AI data centres and ​physical AI, including robotics, Lee’s office said, while the president’s social media posts signalled a new chip cluster planned for the underdeveloped southwest, including Gwangju and South Jeolla province.

Local media have reported the planned investments could exceed 1,000 trillion won ($651.41 billion) over the coming years.

South Korean construction and engineering shares surged in early trade ahead of ​the announcement on expectations that the investments could drive massive regional infrastructure development.

Major cement producers Asia Cement and Hanil Cement jumped 15% and 7%. The ​KOSPI fell more than 2%, with both chipmakers Samsung Electronics and SK Hynix down over 4% and 3% respectively, as global tech stocks took a breather from their ‌recent sharp ⁠rally.

President defends the plan

Besides Samsung Electronics, South Korea is also home to SK Hynix, the world’s two largest memory chipmakers, whose high-bandwidth memory (HBM) chips are pivotal in the global race to build advanced AI systems.

Both companies already operate major semiconductor facilities in and around the Seoul metropolitan area.

The government is also set to detail extensive support measures covering power, water, land, infrastructure, workforce training and housing.

Lee has defended the proposed southwest ​chip hub in a series of ​X posts over the weekend, ⁠rejecting criticism that it favours a liberal stronghold.

He framed it instead as a “national survival strategy” to ease regional imbalances and expand capacity for the AI era.

“The creation of a semiconductor industrial ecosystem in (the southwest) is ​not a special favour for a particular region,” Lee wrote in one post.

“It is the additional creation of ​the most rational ⁠semiconductor industrial centre through the decisions of relevant companies under full government support.”

Industry experts say diversifying chip investment beyond Seoul could ease infrastructure bottlenecks, but warn that building cutting-edge fabs requires vast electricity and water, advanced logistics, deep supplier networks and highly skilled labour - elements that may not scale quickly enough ⁠in a ​new region to meet surging AI demand.

Opposition politicians have sharply criticised the plan, questioning ​whether the proposal is politically motivated, given that 85% of voters in the region backed Lee in last year’s presidential election.

The announcement comes as Lee’s approval rating has slid for ​six weeks to 46.5%, according to pollster Realmeter.

In a letter to EU Technology Commissioner Henna ​Virkkunen released by the Austrian government, Austria’s State ​Secretary for Digitalisation, Alexander Proell, wrote it was ⁠important that Europe was not cut off from major ​innovations.

“Let us jointly explore the strategic establishment and participation of ​Anthropic within the European Union. With legal certainty, market access, capital and a set of values that suit this company,” Proell ​said in the letter.

He did not say how the ​step could be taken and acknowledged there would be scepticism about whether ‌it ⁠was possible.

“The real question is not whether it is easy,” Proell wrote. “The question is whether we Europeans are prepared to be the architects of our technological future, ​or whether we ​wish to ⁠remain mere administrators of decisions made elsewhere.”

Anthropic did not immediately reply to a request ​for comment on the Austrian proposal.

Earlier this month, ​the ⁠European Commission proposed laws to boost domestic cloud, AI and semiconductor industries and cut reliance on US Big Tech, defying ⁠US ​government criticism of the bloc’s crackdown ​on its industry.

The move follows the ‌world-first restrictions that took effect in December, a policy being watched closely by many countries seeking to emulate it, as evidence suggests children are still able to access the ​platforms.

Prime Minister Anthony Albanese said there were still too many children on ​social media and tech firms were not doing enough to comply ⁠with the law.

“We’re calling time on the social media companies today and ​doubling down on the changes that we have made and that we’re prepared ​to make,” he told reporters in Canberra.

“Today, we’ll introduce legislation this afternoon that goes further to ensure social media companies are doing everything within their power to stop children under ​16 from being on their platforms.”

The regulator is investigating possible non-compliance by five platforms: ​Meta’s Facebook and Instagram, Snapchat, TikTok and Google’s YouTube.

There was no immediate comment from Meta, ‌Google ⁠and Snapchat. TikTok declined to comment.

Albanese called on the conservative coalition opposition to back the bill, noting the original policy passed with bipartisan support.

The changes, announced on Sunday, would double fines to A$99 million ($68.2 million) from A$49.5 million.

They also give ​the eSafety Commissioner power ​to compel documents ⁠such as company board minutes and internal emails, ensuring legal cases being built against platforms failing to comply are ​as “strong as possible”, said Communications Minister Anika Wells.

“My message to ​Big Tech ⁠is this: we are not stopping. Every effort you make to frustrate these laws will be met with our efforts to make these laws work,” she said.

“If ⁠the ​eSafety Commissioner finds companies are not doing everything ​they can to comply, they will face the full force of the law.”

Google, owned by Alphabet, told Meta around March ​it could not meet the full Gemini capacity the ​company had sought to purchase, the newspaper said, ⁠adding that the shortfall disrupted and delayed some of ​Meta’s internal AI projects.

Several other Google clients have also been ​affected, though to a lesser extent, according to the report.

Meta has been particularly impacted due to its exceptionally high demand for Google’s ​models, the FT said.

Reuters could not immediately verify the ​report, which cited people familiar with the matter.

Google and Meta did ‌not ⁠immediately respond to requests for comment outside business hours.

Due to the restrictions, Meta has encouraged staff to be more efficient with AI tokens, the units that measure AI usage, ​the FT report ​said.

Even as ⁠companies continue to spend billions on chips and data centres, they are still struggling to ​secure enough computing power to support the growing ​demand ⁠for AI services.

Revenue at Google Cloud grew to $20 billion in the first quarter ended March, but CEO Sundar Pichai said computing ⁠power ​constraints prevented even higher growth and ​contributed to the cloud unit’s backlog nearly doubling quarter on quarter.

More than 100 ‌companies and institutions will now have access to Mythos 5, including many Fortune 500 companies, a source familiar with the new directive said, declining to be identified due to the sensitivity of the matter.

Concern that powerful AI systems could be misused by military intelligence users in China, Russia or other countries of concern has prompted President Donald Trump’s administration to take an aggressive approach to oversight of releases of Anthropic’s and rival OpenAI’s frontier models.

OpenAI said earlier in the day that it ​was delaying a full public launch of GPT-5.6 at the US government’s request, limiting its access to a small group of vetted partners whose details were shared with the authorities.

Anthropic ​had abruptly disabled its most advanced AI models — Mythos 5 and Fable 5 — for all users after the government’s June 12 export control order.

“Today, the government notified ⁠us that Mythos 5, our strongest cybersecurity model, can be redeployed to a set of US organisations that operate and defend critical infrastructure,” Anthropic said in a statement on Friday.

“We’re restoring access for these ​organisations quickly, and we’re continuing to work with the government to expand access to Mythos 5 and make Fable 5 available for general use again,” it added.

Govt criticised for picking who gets access

The government’s vetting ​of which companies can gain access to Mythos has drawn much criticism.

“No one knows how these companies are picked and why everyone else is excluded,” said John Coleman, legislative counsel for the Philadelphia-based nonpartisan free speech organisation, the Foundation for Individual Rights and Expression.

“This is putting too much power in the hands of the government. There’s little transparency, and it raises questions about the rule of law.”

OpenAI boss Sam Altman echoed concerns about the government’s choice of who gets ​access to top models in a post on X.

Extensive safety testing “is not a bad idea. I just don’t like the idea of the government picking the customers,” he wrote.

Experts have said that Mythos models, ​in the wrong hands, could dramatically accelerate sophisticated cyberattacks, particularly in sectors such as banking that rely on complex, interconnected, and often decades-old technology systems.

A letter from US Commerce Secretary Howard Lutnick to Anthropic said there had been “significant ‌progress” in ⁠work done by the company with the government to address “risks associated with the Covered Models.”

It was not immediately clear what safeguards had been adopted.

Anthropic said earlier this month that it understood the government believed there is a method of bypassing, or “jailbreaking,” a safeguard that would prevent Fable 5 from being used in identifying software vulnerabilities.

No export licence needed for non-US citizens

Lutnick said in the letter that an export license will no longer be needed for Mythos 5 to trusted companies and their employees who are not US citizens, or to Anthropic’s employees who are not US citizens, but licensing restrictions will remain in place for ​companies that are not on the approved list.

The source ​said many of the approved companies are ⁠part of Anthropic’s Project Glasswing, which includes about 100 well-known tech companies and institutions.

The government is also moving towards allowing Anthropic to release Fable soon, although a timeline is unclear, the source said.

Both Fable 5 and Mythos use the same underlying AI model, but Fable 5 is designed to be widely available ​for public use, whereas some safeguards are lifted for Mythos.

Both Anthropic and OpenAI plan to go public.

Anthropic’s relationship with the US government has, however, been ​particularly rocky.

The company refused ⁠to allow the US military to use its AI models for domestic surveillance and fully autonomous weapons systems, and the government retaliated by putting it on a national security blacklist.

The government’s restrictions on Anthropic and OpenAI follow Trump’s signing of an executive order this month establishing a voluntary framework for AI developers to offer “covered frontier models” to the US government for up to 30 days before releasing them to trusted partners.

The administration’s latest ⁠order is “a practical ​interim step, but leaves unresolved the larger issue of how companies can widely release updated models,” said Kate Koren, an ​analyst at the Centre for Strategic and International Studies in Washington and a former Commerce Department official.

“The longer there isn’t a system in place that will allow US companies to widely release new models, the more likely it is that China will be ​able to catch up,” she said.

Inclusion in the index typically boosts the stock price, as exchange-traded funds looking to replicate the index’s performance buy shares of the newly ​included firm.

To make it more attractive for companies seeking US listings, Nasdaq, ​along with other index providers FTSE Russell and MSCI, relaxed ⁠its entry requirements, including profitability, the number of days after a company ​goes public and the number of shares available for trading.

SpaceX, which made its ​Nasdaq debut on June 12, has swung between sharp losses and small profits over the past three years. Last year, the company reported a net loss of $4.9 billion.

Large Language ​Model (LLM) makers OpenAI and Anthropic are also expected to file for their ​initial public offerings this year or next year and likely target valuations of more ‌than $1 trillion.

Investors ⁠buy mutual funds and ETFs, such as Invesco’s QQQ and QQQM, that track the Nasdaq 100, to get broader exposure.

J.P. Morgan estimated that SpaceX’s inclusion in the Nasdaq 100 could draw $4.3 billion in passive inflows.

“Clearly, there’s a ​lot of demand; ​that’s why they ⁠fast-tracked the integration into the index,” Michael Field, chief equity market strategist at Morningstar, said. “A lot of people will ​be happy with it.

Some fund managers are less so, the ​sceptics ⁠amongst them, us included. We think the stock is overvalued.“

S&P Global said this month that it was not changing the requirements for SpaceX to enter its ⁠major indices, ​including Wall Street’s benchmark S&P 500 index, ​and will wait for at least 12 months before even considering it.

The strike by Alibaba is described as a “distillation” effort, which Anthropic has said involves training a less capable model on the outputs of a stronger one.

Anthropic said the campaign was conducted between April 22 and June 5, 2026, and generated more ​than 28.8 million exchanges with Claude through almost 25,000 fraudulent accounts.

Anthropic said in the letter that ​distillation is a way to help accelerate China’s ability to reach Anthropic’s advanced Mythos ⁠Preview capabilities.

It said the campaign was conducted by operators affiliated with Alibaba and Alibaba Qwen, Alibaba’s AI lab. ​

Alibaba did not immediately respond to a request for comment.

The letter, dated June 10, was sent to Senators Tim ​Scott and Elizabeth Warren, the chair and ranking member, respectively, of the US Senate Banking Committee, ahead of a scheduled hearing on AI.

In April, the White House accused China of stealing US AI labs’ intellectual property on an industrial scale.

Anthropic said in the ​letter that it was supportive of the US government’s efforts to combat the attacks, including partnering with private sector ​AI companies through threat-intelligence sharing and other exercises.

Anthropic said in a February posting that it had identified a campaign by Chinese AI ‌startup ⁠DeepSeek — whose low-cost AI model sent shockwaves through the technology world in January 2025 — and two other Chinese AI labs to illicitly extract capabilities from its Claude AI platform.

It said DeepSeek’s operation involved over 150,000 exchanges, while Moonshot AI was at a scale of over 3.4 million, and MiniMax over 13 million.

It also said at the ​time that the campaigns were ​growing in “intensity and sophistication” ⁠and that addressing the threat would require “rapid, coordinated action among industry players, policymakers and the global AI community.”

Alibaba was added to the Pentagon’s Chinese military companies list this ​month, a designation it is challenging.

But the Commerce Department has held off placing DeepSeek on ​a trade blacklist, ⁠as Reuters exclusively reported this month, despite it being deemed a national security risk by an interagency governmental committee, as the department tries to avoid escalating tensions with Beijing.

Meanwhile, on June 12, two days after Anthropic sent the ⁠letter, the ​Commerce Department imposed controversial restrictions on Anthropic’s latest Mythos and Fable ​AI models because officials feared they could be deployed by military intelligence users in China and other countries of concern.

The restrictions resulted in ​Anthropic disabling access to the models globally.

The Saudi oil loadings come even though a ship belonging to Taiwan’s Evergreen Marine was hit by an unknown object in the Strait of Hormuz on Thursday.

Middle Eastern producers had been ramping up oil and gas output and exports in the lead-up to the interim deal between ​the United States and Iran to halt the war and reopen the strait, where a fifth of the world’s oil and liquefied natural gas supplies ​used to pass.

Two Very Large Crude Carriers controlled by Saudi’s shipping arm Bahri were seen loading crude at Ras Tanura, the ⁠world’s biggest oil port, while another waited nearby, the data showed. Each VLCC is capable of loading 2 million barrels of oil.

Saudi Aramco, among the last of ​the major Gulf producers to resume exports from inside the Gulf, could not be immediately reached for comment outside office hours.

British Navy agency UKMTO paused its operation to escort ships ​through the strait after the attack on the cargo ship, reigniting concerns about whether the preliminary deal to end the Iran war will hold.

Two US officials told Reuters that Iran had fired on the ship, while Iran’s Persian Gulf Strait Authority, which Tehran established to manage requests for ships to travel through the strait, said vessels outside routes it has set will not be ​guaranteed safe passage.

Ras Tanura port

Ras Tanura sits on Saudi Arabia’s eastern coast on the Gulf and is west of the Strait of Hormuz.

It used to export ​more than 5 million bpd of crude before the conflict.

The country’s largest domestic 550,000 bpd refinery is also located at Ras Tanura, which was shut during the war as a ‌precautionary measure.

Aramco ⁠last loaded a cargo from Ras Tanura port for China on March 8, LSEG data showed, and had to divert its exports to the Red Sea port of Yanbu after the Iranian blockade of the strait during its war with the US, and Israel prevented ships from entering the Gulf.

The war has caused Saudi crude exports to slump to about 4 million bpd in the past three months, the data showed, from more than 7 million bpd in February.

Oil slips on rising supply

Global oil prices ​fell more than $1 a barrel on Friday ​after edging up on reports ⁠of the attack on the cargo ship.

Supply pressure is increasing after crude shipments through the strait rose this week to their highest level since the conflict broke out.

Saudi Aramco may cut August prices sharply next week as competition among producers intensifies.

Iraq’s SOMO and Qatar ​issued tenders offering crude, following similar moves by Kuwait and the United Arab Emirates.

Iran is also rushing its exports ​after Washington temporarily lifted ⁠sanctions.

Two empty VLCCs — Natsumi and Halti — entered the Gulf on Friday to load Iranian oil, shipping data showed.

Tankers ferrying UAE oil continued transiting the strait on Friday, with two laden VLCCs exiting and one heading to Zirku port, the data showed.

“Two million barrels a day came back online in three weeks, and the recovery is spread across the ⁠region,” Rystad Energy’s ​MENA research director Aditya Saraswat said in a note, adding that the supply picture is clearly ​improving.

The consultancy now estimates that shut-in production across the Gulf has fallen to 9.6 million barrels per day (bpd) in mid-June, down from 11.7 million bpd just three weeks ago, and expects a full supply recovery ​in the region by the end of the year.

As pre-orders for the year’s biggest game open on Thursday, here is what we know about the latest instalment in the juggernaut series that has sold over 470 million copies worldwide.

What’s the plot?

Every “Grand Theft Auto” title since the 1997 original has put players in the shoes of a gangster working up the ladder of organised crime in a fictional American metropolis, with the backdrop changing for each episode.

This time, the protagonists are a Bonnie and Clyde-style couple of hoodlums, Jason Duval and Lucia Caminos.

The game’s website says the pair “find themselves… in the middle of a criminal conspiracy” as the plot unfolds.

Other members of the larger-than-life cast include a conspiracy-theorising hacker, a strip club owner and two women rappers-influencers.

Images from the two trailers Rockstar has released so far feature breakneck car chases, scantily clad women and innumerable winks to social media culture.

Where is it set?

The plot of “GTA VI” takes place in the fictional US state of Leonida, a stand-in for Florida, and its megalopolis Vice City, the Miami-inspired setting for a “GTA” title of the same name released in 2002.

So far, the published imagery shows off a highly detailed urban world of varying neighbourhoods teeming with non-player characters (NPCs).

Protagonists are shown driving cars, shopping in a grocery store and working out with weights at the beach.

Players will also be able to explore beyond the city, including in a swampy region reminiscent of Florida’s Everglades and a tropical archipelago.

What is there to do?

Heists and gunfights will once again take pride of place in the game’s mechanics.

As in previous episodes of the “open world” series, players will be able to take on missions that move the main story forward, ranging from robberies to murder and drug trafficking.

But one of the major draws of “GTA” is the freedom for players to simply explore the world — whether abiding by the law or causing as much mayhem as possible while dodging authorities’ increasingly muscular efforts to bring them to justice.

On offer will be vehicles ranging from motorbikes and cars to jetskis and aircraft.

What’s new in the sixth instalment?

On its website, Rockstar Games calls the coming title “the deepest and most immersive GTA experience yet”.

Available only for the latest-generation Playstation 5 and Xbox Series consoles, it promises far higher graphical fidelity and detail compared with its 13-year-old predecessor, which was designed for much less capable hardware.

Players will likely experience far shorter loading times when travelling between different areas of the game.

It is also the first time since 1997 that the series has put a woman in a starring role, after years of criticism of Rockstar for its often crude depictions of female characters.

The many set pieces inspired by viral videos shown off in the trailer hint that “GTA VI” will pick up the satirical baton from its predecessors, taking aim at the domination of social networks in the public sphere.

When is it released and what does it cost?

Unless delayed for a third time, “GTA VI” will be available to play on November 19 in digital or physical editions.

Rockstar says the versions available to buy in stores will have a download code inside the box to install the game on a console.

In the United States, “GTA VI” will cost $79.99 for the standard version and $99.99 for an “Ultimate” edition with bonus content such as additional weapons and vehicles.

Players who have pre-ordered the title will be able to download it onto their console from November 12, but can only launch the game on the official release date.

New Delhi has largely blocked Chinese companies from entering the market since 2020, and now Beijing is clamping down on the export of its ‌tech know-how.

Yet ties between the two countries’ carmaking industry are only growing.

Tata Motors said earlier in June it would use Chery’s carmaking platform to manufacture premium EVs in India.

The deal doesn’t involve an equity stake, and both companies stressed it is a supply arrangement without any transfer of technology know-how to Tata, highlighting the political sensitivities.

India ramped up scrutiny of Chinese businesses after a 2020 border clash between the two countries killed soldiers on both sides.

While New Delhi ​and Beijing are working to improve ties, some friction remains.

“If India wants to expand its manufacturing sector and be a bigger part of the global supply chain, a partnership with China is inevitable. ​

If Chinese companies want to be global leaders, they cannot wish away India and its economic potential,“ said Santosh Pai, partner at law firm Dentons ⁠Link Legal.

For Tata, India’s third-largest automaker, Chery’s platform offers a quicker way to launch EVs.

Tata plans to eventually shift from relying on imported kits from China to developing components locally — a move seen ​favourably by some Indian policymakers because it would boost Indian manufacturing.

“We are supportive of deals that lead to more local manufacturing or supply-chain shifts down the road. That is a good way to approach China,” ​said a senior Indian government official.

For Chinese carmakers grappling with a slowdown at home and excess manufacturing capacity, such deals could be the answer to boosting revenue without violating Beijing’s export control orders.

Tata and Chery did not respond to requests for comment.

Growing market

The Tata-Chery deal shows that, despite its best efforts, India can’t keep China’s EV industry completely out.

The world’s most advanced EV industry is likely to continue to make inroads into India, a huge and still growing market.

That’s bad ​news for Japanese automakers and others who are investing big in India – in part because they don’t face major competition from Chinese rivals there now.

Chinese EV makers understand the importance of gaining a foothold ​in India through such supply deals, said Gao Hua, a former director at China SAE and now an independent analyst.

“If Chinese firms don’t participate, others from different countries will step in,” Gao said.

Chinese partnerships are increasingly appearing in ‌sectors long ⁠dominated by Japanese, Korean and European firms, and they are challenging the incumbents with technologies that many analysts say are cheaper and faster to deploy.

For instance, Indian component maker Uno Minda has a joint venture with China’s Inovance to manufacture EV powertrains in India - a sector where Bosch, Nidec and Aptiv are already present.

Battery co-operation halted

Technology licensing deals between India and China started to gain traction in the aftermath of the 2020 investment restrictions.

But it wasn’t all smooth sailing. In 2025, Beijing’s export control curbs in retaliation to Trump’s tariffs forced Indian battery maker Amara Raja to end its licensing deal with China’s Gotion for ​lithium-ion cell technology for EV batteries.

“All technical collaboration ​has stopped,” Amara Raja’s executive director Vikramadithya Gourineni ⁠told Reuters.

“The main things we were able to take away were understanding of factory and line layouts, technology roadmaps … and connecting to the vendor base,” Gourineni said.

Because the licensing deal was no longer possible, Amara Raja is instead ramping up investment in in-house R&D and talent, he said.

The company is now importing ​equipment, battery cells and other materials from Chinese suppliers to meet its cell manufacturing ambitions, but it struggles to get enough visas for engineers ​to come from China for ⁠operational support.

Chery’s other Indian partner

Last year, steel-to-cement billionaire Sajjan Jindal’s maiden carmaking venture, JSW Motor, agreed to a partnership with Chery similar to Tata’s.

Under the deal, JSW has secured rights to use and adapt multiple Chery platforms to build a range of hybrids and EVs for India, sources familiar with the plans told Reuters.

This involves an upfront payment of about 20 billion rupees ($209 million) plus royalties, one of the people added.

JSW, which ⁠is investing $3 billion ​in the venture, is targeting sales of 300,000 vehicles by 2030, the sources said.

The initial vehicles will largely come as imported ​kits from Chery, with JSW gradually building out an Indian supply chain and scaling up car production at its factory in western India, they added.

JSW Motor and Chery did not respond to requests for comment.

“This highlights the importance of nuanced approaches. Cutting ​ties is not always the best option,” Gao said.

The ransomware group has posted more than 200,000 files on the dark web, the security researchers told Reuters.

“A few weeks ago, Tata Electronics identified a cybersecurity incident on some of our systems.

Our response protocols were deployed immediately, and the incident has had no impact on our operations across businesses, which remain unaffected,“ Tata Electronics told Reuters in a statement.

Report Ad

Apple was investigating the breach, and ​a “full analysis was going on”, a source familiar with the matter said, adding that Tata had received a ransom demand related to the incident.

Apple did not ​respond to requests for comment. Tata Electronics declined to comment on the ransom demand.

The breach is the latest setback for Apple’s supply ⁠chain in India, where Tata faces scrutiny over alleged contamination of farmlands near one of its iPhone parts plants, Reuters reported.

Tata is emerging as one of Apple’s most important manufacturing ​partners outside China, an expansion that is a cornerstone of Prime Minister Narendra Modi’s push to make India an electronics manufacturing powerhouse.

Tata was hit by a cyberattack on its British ​Jaguar Land Rover group last year that resulted in a six-week output halt.

The Indian Computer Emergency Response Team, a unit under India’s IT ministry that oversees cyber incidents, did not immediately respond to Reuters emails seeking comment.

Apple’s factory data’

World Leaks, which has previously claimed responsibility for a Nike break-in, said on its dark net website that it was publishing stolen data from Tata Electronics.

Reuters could not immediately verify the authenticity of ​the data and could not immediately reach World Leaks for comment.

The World Leaks website says the Tata Electronics data comprises more than 200,000 files totalling over 630 gigabytes. ​

A database on its website shows several purported Apple files and folders, some titled “com.apple.factorydata”, and documents referring to “material specification”.

Indian cybersecurity researcher Rajshekhar Rajaharia, who reviewed the Tata files on World Leaks for Reuters, ‌said they ⁠also contain emails, event logs spanning several years and passport copies of employees, including foreign nationals.

Rajaharia has previously advised the Indian police on cyber incidents.

The website is only accessible on the dark web, or dark net, beyond the reach of search engines.

A second security researcher who reviewed the data dump, Rakesh Krishnan, told Reuters it had been accessible on the dark web since at least June 10.

Tesla documents, ‘Trade secret’

Tata also makes parts for Tesla, according to industry sources.

One folder on the World Leaks database was labelled “NV36 Chargeport Controller - North America”, a ​purported reference to parts used in an ​upgraded version of Tesla’s Model Y SUV.

Another ⁠purported Tesla 2023 document described as “TRADE SECRET” showed certain drawings for its project Highland - a publicly known internal codename for its revamped Model 3 sedan.

Tesla did not respond to requests for comment.

Rajaharia also shared a screen recording of his review of the files.

It showed ​a search for “Apple” returned 181 files and folders, while a search for “Tesla” returned files including what appeared to be manufacturing specifications ​and an assembly document ⁠dated May 2025.

Some files published by World Leaks carried footers saying, “This document contains proprietary and confidential information of Apple Inc.” and “information contained herein is deemed confidential, proprietary, and a trade secret of Tesla Inc.”

The breach highlights the vulnerability of global businesses to increasingly sophisticated cyber and ransom attacks.

Among the files was a 52-page document bearing Apple’s proprietary markings purportedly detailing quality inspection ⁠standards for iPhone ​circuit board components.

There were also 33 files and folders for the search term “Hosur” - the location of Tata’s ​main iPhone assembly plant in Tamil Nadu state.

Tata informed some employees at its iPhone assembly operations last week of the data breach, said a second industry source familiar with the matter.

Tata currently accounts for roughly a ​third of Apple’s iPhone production in India, with Foxconn making up the rest.

Across the world, hiring is getting faster — but not necessarily smarter.

According to an article in Khaleej Times, AI-powered recruitment platforms can now screen thousands of profiles in minutes.

Interviews are increasingly automated. Talent pools are expanding globally.

Although the system appears efficient on the surface. Beneath that speed lies a much deeper challenge: trust.

Multiple global surveys indicate that a substantial proportion of CVs include exaggerated or misleading information.

At the same time, employers are reporting growing difficulty in verifying credentials, particularly in cross-border hiring environments.

As recruitment becomes more algorithmic, the quality of input data is becoming critical.

AI can match talent — but it cannot independently confirm whether that talent is real. That is the structural gap emerging today.

From static CVs to verifiable career records

The traditional CV was designed for a paper-based world.

It is static, self-declared, and only updated when a candidate chooses to revise it.

It rarely shows whether achievements were verified, whether skills were demonstrated, or whether credentials are authentic.

Yet most AI recruitment systems still treat it as structured truth.

If the foundation is unreliable, then the output — no matter how advanced the algorithm — will also be unreliable.

What is beginning to emerge instead is a different model: the CV as a live, verifiable professional record.

A living profile that updates as someone gains new certifications, completes roles, earns promotions, or proves new skills — where every step is backed by verified, trusted credentials.

This shift is not only technological. This is philosophical as well. It reframes hiring from self-declaration toward verified merit.

Skills validation is still evolving

Across the world, employers are caring less about where someone studied and more about what they can actually do.

The UAE has become a hub for cutting-edge technologies like AI, fintech, cybersecurity, and digital innovation.

As a result, the question is no longer whether talent exists — but whether that talent can clearly prove its skills in a way employers trust.

Competency-based certification models are gaining momentum worldwide.

Instead of broad academic transcripts, professionals demonstrate applied skills through projects, assessments, and measurable outputs, the report said.

The value lies not in the certificate itself, but in the verification of what that certificate represents.

When future-skills training is directly tied to measurable capability, the hiring equation changes.

Employers gain clearer signals. Candidates gain stronger credibility. AI systems receive higher-quality data.

In building emerging-technology certification programmes such as Certified Emerging Technologies Analyst (CETA), we have observed that employers respond differently when assessments are skills-based and digitally verifiable rather than descriptive.

The focus shifts from “What did you study?” to “What can you deliver?”

AI matching depends on verified inputs

AI-powered talent platforms promise precision —matching job descriptions with candidate capabilities in seconds.

However, research in HR technology adoption consistently shows that mis-hires remain among the most expensive organisational mistakes, often driven by incomplete or inaccurate profile data.

When job requirements are precise and candidate records are verified, AI matching becomes genuinely transformative.

It can reduce bias, shorten hiring cycles, and expand access to cross-border talent.

But when inputs are unverified, automation simply scales inefficiency.

This is why infrastructure matters more than interfaces.

In developing workforce platforms such as TruCV and TruJobs at Edubuk, experience has reinforced a simple point: AI is not the starting point. Trust is.

Once credentials are verifiable and professional records are tamper-resistant, intelligent matching becomes not only possible — but meaningful.

The real innovation lies not in dashboards or predictive scoring, but in systems where every skill, certification, and experience can be validated without friction.

The Middle East opportunity

The UAE and the wider Middle East have a distinctive advantage.

Unlike legacy-heavy markets constrained by older HR systems, the region is digitally ambitious. Governments are proactive. Enterprises are agile. Cross-border talent mobility is high. This creates space to lead.

By embedding verification into the core of digital hiring, the region can move beyond adopting AI tools and instead begin architecting a trusted workforce infrastructure.

The distinction is important: one optimises processes, the other reshapes standards.

Globally, several forces are converging — AI recruitment systems, digital credentials, decentralised verification technologies, and skills-based education reform.

The regions that integrate these layers effectively will help define the future of work.

Beyond efficiency, toward integrity

Hiring has always been an exercise in reducing uncertainty.

Technology has made it faster. Now it must make it more reliable.

The next evolution in workforce systems will not be measured by how many resumes are processed per second, but by how confidently organisations can say the data they rely on is authentic.

AI can match talent. But only trusted infrastructure can ensure that talent is real. The future of hiring is not simply artificial intelligence — it is verifiable intelligence.

Facing a broad decline in education test scores, the government in 2024 banned smartphones from schools and gave teachers back more powers to enforce discipline in the ​classroom.

Using AI increases the risk that young children skip important ⁠steps in their education, Prime Minister Jonas Gahr Stoere told a press ​conference on Friday.

“The most important thing in school is that our children ​learn to read, write and do mathematics,” Stoere said, adding that the new standards will be imposed from the new school year beginning in late August.

Pupils from first ​through seventh grade, aged 6 to 13, should, as a general ​rule, not be using AI, while those in lower secondary school, aged 14 to ‌16, can ⁠cautiously adopt tools under teachers’ supervision, the government said.

In upper secondary education, from ages 17 to 19, students should learn to use AI appropriately so that they are prepared for further education and work, it added.

Norway ​began adopting computers ​in classrooms in ⁠the 1990s and tablets after the introduction of the iPad from 2010 onwards, reducing the reliance on books ​and handwriting.

But in a related statement on Friday, the ​government also ⁠said it will propose legislation to fund the use of more books in classrooms, reversing the trend towards computer tablets.

The Norwegian government in April also ⁠announced ​plans to ban children from using social media ​until they turn 16, following a trend pioneered by Australia and some other nations to reduce young ​people’s use of electronic devices.

Senior Anthropic technical staff were scheduled to meet with Trump administration officials earlier this week to discuss a ​dispute over foreign access to its most advanced AI ​models, Fable 5 and Mythos 5.

The company last ⁠week disabled access for all users to those models after ​Trump ordered Anthropic to block foreign nationals from accessing them.

Here are ​some of the details from the Axios interview:

When asked if he viewed Anthropic, or its CEO Dario Amodei, as a threat to national security, ​Trump said: “Well, not now, but a week ago, maybe.”

Trump told ​Axios that Amodei responded to the administration’s export control directive “very quickly” and “responsibly.”

Trump and ‌other ⁠G7 leaders met with tech bosses, including Amodei, at a summit in France this week.

Trump did not rule out using emergency powers under the Defence Production Act against Anthropic, according to ​Axios.

“I have the ​power to use ⁠a lot of things,” Trump said of the DPA. “But I’m not sure I have to do ​that.”

Asked to comment on Trump’s interview, an Anthropic ​spokesperson ⁠said: “We are grateful to the administration for their ongoing partnership in working to get this matter resolved as quickly as possible.

We ⁠remain ​committed to working alongside them towards our ​shared goals of protecting critical infrastructure and making sure the US leads in ​AI.”

The Australian carrier, which plans to follow up with non-stop flights from Sydney to New York, this week offered detailed briefings ‌on the science of roughly 20-hour journeys as it looks to convince passengers to pay a premium for avoiding a stopover.

“It’s a major biological challenge crossing all these time zones - seven to nine for London and 14 to 16 for New York,” said Peter Cistulli, professor of sleep medicine at the University of Sydney, who took part in scientific research for “Project Sunrise”.

Frequent long-haul Australian travellers told Reuters their biggest priorities when deciding whether they would take the ultra-long flights included seat comfort, ​the ability to move around and cost.

Qantas has gone beyond that, studying every detail for the flights on Airbus A350-1000ULR planes since the project was launched nearly a decade ago, from nutrition to ​ergonomics, movement and above all, light - a vital cue for the body clock.

By optimising meal times - such as avoiding food just after takeoff - and creating a “protective ⁠sleep window” with the help of lighting, passengers showed better alertness during tests than with a traditional service, Cistulli said.

Cabin designer David Caon said he had been asked to approach it as a health and ​scientific task as much as an aesthetic one.

“When you have a passenger for essentially a whole day, it really does drive a whole set of new decisions,” he said.

Caon toyed with novel ideas, including exercise bikes ​and yoga mats. Those did not make the cut, but a dedicated “wellness zone” remained, bathed in diffused, shimmering light.

“I wanted to recreate the sense of lying by the swimming pool,” Caon told reporters.

Elsewhere in the cabin, custom mood lighting will simulate sunrise or sunset, moving from front to back. It took weeks to programme 14 light “scenarios” inspired by Australia’s scenery.

All of this aims to reduce the discomfort of flights that could be as long as 22 hours on such specially modified ​jets.

But the sense of space also has a business aim: to turn Australia’s remote position into a travel experience that other airlines cannot easily match - and generate 20% more revenue.

Qantas CEO Vanessa Hudson said the ​expected premium over one-stop flights was based on results seen between Perth and Europe.

Most analysts say the performance of its 17-hour Perth-London flights bodes well for the Project Sunrise business case.

PASSENGER VIEWS

Qantas has designed the plane in a ‌premium-heavy configuration ⁠as it will need to squeeze all the profit out of just 238 passengers due to weight restrictions.

It may have to block some seats to save fuel when the weather dictates.

A senior airline industry executive said other operating risks would include costly diversions, given the long routes.

Sam Davies, who works in the drinks marketing industry, said he already uses the Perth flights on his way to Sydney from his home in Paris and would look at the non-stop option.

“There is something wonderful about waking up in Australia and not having to get off anywhere and go through security and kill three hours, so I am all up for ​it,” he said.

But comfort would be decisive.

“I am six-foot-four (193 ​cm), so the economy seat is too small…I ⁠would have to ask for some more details on the seats,” Davies added.

Qantas said the regular economy pitch, or distance between seats, would be 33 inches, though some rows would be slightly tighter at 32 inches, and this would be made clear when booking.

Part of the cabin will be sold as “Economy ​Plus,” offering 34 inches of legroom.

At the front of the aircraft, Qantas is the latest airline to install enclosed first-class suites with a fixed bed.

Melbourne-based business ​executive Ian Morden said he ⁠was not put off by ultra-long flights, as he likes to use the time to work and think.

But he questioned how much the four-hour time savings from avoiding a stopover would justify the fares that the airline’s business case demands.

“A slight premium would be justifiable, but…I probably wouldn’t choose it for a 20% premium on an already much more expensive business-class flight,” he said.

London-based Nathalie Curtis, who travels extensively for her work in the international cultural ⁠sector, said ​she would take the flight if it lived up to the airline’s marketing, but expressed concerns about the deterioration of cabins ​on very long flights.

“If it allows you to move around, reduce jet lag with lighting adjustment and is hygienic and saves … four hours, then I would go for it and pay a 20% premium,” she said.

Mark Levine, an Australian strategic adviser in New York, ​said direct flights from Sydney would take the organisational friction out of lives spread between different continents.

“The distance doesn’t change, but the journey feels a little smaller,” he said.

SNAP-8 represents a truncated fragment engineered to mimic specific domains of SNAP-25, thereby enabling targeted interference with protein-protein interactions that govern vesicular docking and release. Research indicates that such peptides might serve as valuable molecular probes for studying regulated exocytosis, neurotransmitter release analogue systems, and intracellular trafficking pathways.

Rather than functioning as a direct substitute for endogenous proteins, SNAP-8 is believed to operate as a competitive modulator within SNARE-mediated processes. This conceptual positioning has led to increasing interest in its application across multiple research domains, including neurochemical signalling, cellular communication, and membrane biophysics.

Structural characteristics and molecular design

SNAP-8 is composed of eight amino acids designed to replicate a functional motif of SNAP-25. This region is theorised to participate in the formation of the SNARE complex, which typically includes syntaxin, synaptobrevin (VAMP), and SNAP-25 itself. The assembly of this complex is thought to facilitate the close apposition of vesicular and target membranes, ultimately enabling membrane fusion.

The truncated nature of SNAP-8 suggests that it may not fully replicate the structural complexity of the parent protein. However, investigations purport that its minimalistic design might allow for selective interaction with specific SNARE components. This selective binding may disrupt or attenuate the formation of functional SNARE complexes, thereby influencing vesicular release dynamics.

From a biochemical perspective, the peptide is thought to exhibit enhanced stability relative to longer protein fragments due to its reduced structural complexity. Additionally, synthetic modification strategies have been explored to improve its resistance to enzymatic degradation in controlled research environments, although such modifications remain context-dependent.

Interaction with SNARE complex machinery

The SNARE complex represents a highly conserved molecular system responsible for vesicle fusion across diverse cellular contexts. SNAP-25 seems to contribute two α-helical domains that participate in the formation of a four-helix bundle, a structure essential for membrane fusion. SNAP-8, as a fragment derived from this protein, is hypothesised to interact with similar binding interfaces.

Research indicates that SNAP-8 might compete with endogenous SNAP-25 for binding to syntaxin or other SNARE-associated proteins. This competitive interaction may alter the kinetics of SNARE complex assembly, potentially slowing or modulating vesicular fusion events. Such modulation has been of particular interest in experimental systems investigating neurotransmitter-like release mechanisms, even outside classical neuronal contexts.

Implications for neurochemical signalling research

Although SNAP-8 is not a neurotransmitter itself, its interaction with SNARE machinery places it at the intersection of neurochemical signalling research. Vesicular release mechanisms underpin communication in many specialised cellular systems, and peptides that may influence these processes have been hypothesised to provide insight into how signalling precision is achieved.

Investigations suggest that SNAP-8 might be utilised in experimental frameworks to explore how vesicle release frequency, timing, and amplitude are regulated. Studies suggest that by introducing a controlled perturbation into the SNARE system, researchers may observe shifts in signalling patterns that reveal underlying regulatory principles.

Possible applications in cellular communication and secretion models

Beyond neurochemical systems, vesicular trafficking plays a critical role in many forms of cellular communication, including hormone-like secretion, immune signalling analogues, and intracellular transport. SNAP-8 has been explored as a tool to probe these processes in controlled laboratory settings.

Research indicates that the peptide might influence the release of signalling molecules packaged within vesicles, thereby altering communication between cells in research models. This modulation may provide a window into how secretion is coordinated and how disruptions in vesicular dynamics might impact broader system behaviour.

Considerations in experimental contexts

While SNAP-8 appears to offer a range of research applications, its use requires careful consideration of experimental parameters. Research indicates that the peptide’s interaction with SNARE components may vary depending on the composition of the system, the presence of accessory proteins, and the specific conditions under which experiments are conducted.

Research indicates that concentration-dependent dynamics may play a role in determining the extent of SNARE modulation. At different levels, SNAP-8 appears to exert varying degrees of influence on vesicular processes, necessitating precise calibration in experimental design.

Future directions and theoretical perspectives

The continued exploration of SNAP-8 is likely to intersect with advances in structural biology, computational modelling, and high-resolution imaging techniques. These approaches may provide deeper insight into how the peptide interacts with SNARE components at the molecular level.

It has been theorised that future research may focus on refining SNAP-8 analogues with enhanced specificity or altered binding characteristics. Such modifications could expand its utility as a research tool and enable more precise interrogation of vesicular dynamics.

Conclusion

SNAP-8 represents a compelling example of how synthetic peptides may be derived from functional protein domains to serve as targeted modulators in experimental research. By interacting with the SNARE complex, this octapeptide has been theorised to influence vesicular fusion processes, offering insights into the mechanisms that govern intracellular communication. For more useful peptide data, visit Biotech Peptides.

References

[i] Rizo, J., & Xu, J. (2015). The synaptic vesicle release machinery. Annual Review of Biophysics, 44, 339–367. https://doi.org/10.1146/annurev-biophys-060414-034057

[ii] Chen, Y. A., & Scheller, R. H. (2001). SNARE-mediated membrane fusion. Nature Reviews Molecular Cell Biology, 2(2), 98–106. https://doi.org/10.1038/35052017

[iii] Südhof, T. C., & Rothman, J. E. (2009). Membrane fusion: Grappling with SNARE and SM proteins. Science, 323(5913), 474–477. https://doi.org/10.1126/science.1161748

[iv] Südhof, T. C., & Rothman, J. E. (2009). Membrane fusion: Grappling with SNARE and SM proteins. Science, 323(5913), 474–477. https://doi.org/10.1126/science.1161748

[v] Jahn, R., & Scheller, R. H. (2006). SNAREs—engines for membrane fusion. Nature Reviews Molecular Cell Biology, 7(9), 631–643. https://doi.org/10.1038/nrm2002

The company has unveiled “Midjourney Medical,” which is developing a scanner that can reportedly generate detailed 3D images of muscles, organs, bones and body composition in around 60 seconds while a user stands in a shallow pool of water inside a spa-like facility.

Unlike MRI or CT scans, the system uses multiple ultrasound sensors arranged in a ring around the body.

These sensors emit sound waves from different angles, with computer systems reconstructing the data into full-body images.

The company says the process is non-invasive, fast and does not use radiation.

The move marks Midjourney’s first step into hardware and healthcare, expanding beyond its core business in AI-generated imagery.

Founder David Holz said the long-term goal is to make preventive health scanning as routine as dental check-ups.

The company plans to build up to 50,000 scanners capable of processing up to one billion scans per month by 2031, starting with a wellness centre in San Francisco expected to open in 2027.

Initially, the scanner will be positioned as a wellness product rather than a medical diagnostic tool, with regulatory approval being sought for broader clinical use in the future.

The announcement has drawn interest from parts of the technology sector, but medical experts have urged caution about how the scans are interpreted.

According to reports, radiologists warn that widespread screening could lead to false positives, potentially causing unnecessary anxiety and follow-up procedures.

They also note that ultrasound imaging has limitations compared to MRI and CT scans in diagnosing certain conditions.

Cardiologist Eric Topol told MarketWatch the technology shows promise but is unlikely to replace MRI for applications such as brain imaging, suggesting it may instead complement existing preventive screening methods if supported by clinical evidence.

Science communicator Hank Green also said ultrasound should not be directly compared with MRI, stressing that it remains one of several diagnostic tools rather than a replacement.

The company says the system primarily relies on ultrasound hardware and computational imaging, with artificial intelligence expected to play a larger role later in analysing scan data once regulatory approvals are secured.

The announcement marks the latest step in the marketing campaign for the next instalment of the blockbuster Grand Theft Auto franchise, one of the most commercially successful entertainment properties in history. Rockstar shared the artwork and confirmed that players will be able to pre-order the game beginning next week.

The company has not yet announced pricing details. Pre-orders will be available through the PlayStation Store, Xbox Store and participating retailers.

Rockstar and parent company Take-Two Interactive have repeatedly reaffirmed the Nov. 19, 2026, release date after previous delays pushed the game beyond its original launch window.

The game is set in the fictional state of Leonida, inspired by Florida, and follows protagonists Lucia Caminos and Jason Duval in a modern-day return to Vice City.

Rockstar has not announced a PC release date. The company has only confirmed launches for PlayStation 5 and Xbox Series X|S, while a PC version remains unannounced.

Grand Theft Auto VI is scheduled to launch on Nov. 19, 2026, for the PlayStation 5 and Xbox Series X|S.

The resolution prohibits children under 15 from ​creating ​or ⁠using personal social media ​accounts and ​restricts ⁠their access to the platforms’ ⁠full ​features.

DeepSeek, CXMT and other companies were approved by an interagency committee last year for addition to the Commerce Department’s Entity List, which is being reported for the first time.

Reuters is also exclusively reporting the large number of companies awaiting publication on the list.

DeepSeek, whose low-cost AI model sent shockwaves through the technology world in January 2025, has supported China’s military and intelligence operations, a senior US State Department official told Reuters last ​year, adding that the startup tried to use Southeast Asian shell companies to illegally access advanced US chips.

This year, Anthropic said it identified a campaign by DeepSeek and two other Chinese AI ​labs to illicitly extract capabilities from its Claude AI platform to improve their own models, and OpenAI warned lawmakers that DeepSeek also was targeting its models.

ChangXin Memory Technologies, ⁠China’s top memory chipmaker, was designated as a Chinese military company by the Defence Department under the Biden administration.

The Commerce Department considered placing it on its Entity List more than a year ago, Reuters ​and others reported.

U.S. companies cannot ship goods, software and technology to companies on the list without a license, which is likely to be denied.

DeepSeek and CXMT could not be reached for comment outside normal business hours.

​The Commerce Department’s Bureau of Industry and Security, which oversees the list, did not directly respond to questions about why updates to the Entity List had not been published since last year, or comment on DeepSeek and CXMT.

The bureau uses “many policy and enforcement tools, including the Entity List … daily to ensure we are combating bad actors,” BIS said in a statement.

Tense rivalry

The United States and China are locked in a tense rivalry over technology, trade and national security, with Washington ​using tariffs and export controls to keep Beijing at bay while China maintains a stranglehold on rare earth minerals that defence, auto and chipmaking firms need.

The US has not posted any additions to its Entity ​List since October, the longest stretch between new postings in more than a decade, said Philip Luck, who studies global supply chains at the Washington-based Centre for Strategic and International Studies.

“The Entity List is like whack-a-mole, and you’ve got to ‌keep whacking ⁠the moles,” Luck said, referring to an arcade game.

The lack of new listings is likely allowing American technology to reach adversaries who could use it against the US, he added.

“The fact that the US hasn’t put any companies on the Entity List since October demonstrates that trade policy is overshadowing the use of a critical national security tool,” said Kevin Kurland, a former Commerce Department official.

Multiple Chinese companies were slated for the list for supplying Russian drones that were recovered in Poland last September, one of the people said.

Listing those lesser-known companies is even more important to US suppliers who may not know the nature of their business, the person ​said.

Dozens of other Chinese companies were identified last year ​as national security risks for selling restricted Nvidia ⁠chips to Chinese universities, but were not added to the list, a third source said.

Chinese companies that make and sell drones and robot dogs for the country’s military were also selected as potential targets, according to the third person.

Since late 2025, Jeffrey Kessler, under secretary of commerce for industry and security, has sought to ​avoid listing Chinese parties for fear of escalating tensions between the US and China, according to the first source and other people familiar with the ​matter.

The dearth of listings offers ⁠a window into what many see as a larger problem at the Bureau of Industry and Security under the second Trump administration — an inability to act or issue new rules to combat threats that can be reduced by restricting exports.

Early last year, for instance, the bureau said it would replace a regulation created under former President Joe Biden to govern global access to US-origin AI chips.

But it has still not published a replacement, and is not enforcing ⁠the earlier rule, opening ​a potential loophole that may have allowed the chips to be exported to Chinese companies outside China.

Decisions regarding whether to add ​an entity to the list are made by an interagency committee, which includes officials from the departments of Commerce, Defence, Energy, State and sometimes Treasury.

,But the first two sources said the committee has approved companies for the list, and Commerce has not published ​them.

At least 75 Chinese entities in advanced semiconductor production, semiconductor manufacturing equipment production and AI modelling have gone through the committee and were slated for blacklisting, one of the sources said.

SpaceX has announced plans to acquire AI coding startup Cursor in an all-stock deal valued at $60 billion, expected to close by the third quarter of 2026.

The deal would significantly increase the wealth of Cursor’s four cofounders — Michael Truell, Aman Sanger, Sualeh Asif and Arvid Lunnemark — each of whom is projected to become a billionaire worth about $2.7 billion.

Early investors, including Andreessen Horowitz and Thrive, are also set for substantial returns from the transaction.

Cursor, founded in 2022, began as a developer-focused coding tool but has since expanded rapidly into AI-driven programming assistance.

The company reportedly crossed $4 billion in annualised revenue earlier this year, driven by its “Cloud Agents” product, which handles complex programming tasks over extended periods.

The startup had previously faced increased competition from major AI firms, but recent growth has strengthened its position in the sector.

SpaceX and Cursor first established a partnership earlier this year, with both companies collaborating on AI model development using large-scale computing infrastructure.

SpaceX has been expanding its AI ambitions in recent months, including deeper investment in computing capacity and partnerships aimed at advancing generative AI systems.

No regulatory approvals or closing conditions for the acquisition have been disclosed yet.

Only after Rossella’s death did Irene ​and her husband unlock her devices.

They found that she had been using social media far more than they had known, including maintaining a secret Instagram profile called ‘Just a dead person’ with a zero instead of an ‌o.

In September 2023, they said, Rossella began searching for depressive material which mirrored how she felt. Social media algorithms kept pushing it back to her, and just five months later, she was dead.

“At some point, it seemed to take on a life of its own, growing until it overwhelmed the cheerful, sociable side of her — the brighter part,” Irene told Reuters in a private room at a café in central Asti, her hometown in northern Italy.

Rossella’s parents are among several families in Italy that have brought a lawsuit against Instagram-and-Facebook-owner Meta, and its biggest social media rival TikTok.

In the first collective action in Italy to directly challenge ​social media companies and their algorithms, the families are seeking tighter limits on minors’ access and greater awareness of risks.

Both companies deny the lawsuit’s allegations that their services are harmful to young people, and say they take steps to protect young ​users by removing harmful content, limiting exposure to risky material and helping families manage children’s accounts.

“We know parents worry about the safety of their teens online, which is why we’re consistently making changes to ⁠help protect teens,” a Meta spokesperson said, citing its “Teen Accounts” and built-in safeguards.

“We strongly disagree with these allegations, which ignore our longstanding commitment to supporting young people.”

TikTok said its efforts include strict enforcement of guidelines aimed at protecting users’ mental and behavioural health, adding that ​it removes more than 99% of content that violates those rules.

“We also continue to invest in safety measures to diversify recommended content, block potentially harmful searches and connect vulnerable users with support resources,” a TikTok spokesperson said, citing local suicide prevention help lines.

Asked specifically about the role ​that Instagram may have played in Rossella’s case, Meta told Reuters that it would not comment directly during the litigation, but that young people’s mental health is shaped by a wide range of factors.

The impact of social media platforms depends on how they are used, the safeguards in place to protect users, and the level of parental involvement, it said.

A sudden illness

Speaking slowly and choosing her words carefully, Irene said Rossella’s tragedy unfolded like a sudden, devastating “illness” that left her parents powerless.

Without the algorithm, she says, “the progression of her distress — or psychosis, or whatever it was that I still cannot define — might have unfolded more naturally.”

Scrutiny ​of digital platforms is intensifying across Europe, with Britain announcing plans this week to ban social media for children under 16.

In the United States, a US ruling found Meta and Alphabet’s Google negligent in designing platforms deemed harmful to young people.

European Union regulators are stepping up enforcement ​of the Digital Services Act, pressing online platforms to better protect minors and curb harmful content.

“The goal is not to dismiss the benefits of social media, but to remove the technological and marketing mechanisms that make it harmful to the most vulnerable users,” said lawyer Stefano Commodo, who is leading ‌the case with ⁠the Italian association of parents MOIGE.

Parents cannot keep up: The limits of control

Parents say safeguards provided by the platforms fall short, noting that children can easily find online tutorials showing how to bypass filters or avoid time limits by switching devices.

“Monitoring social media use is a full-time job. It would require parents to spend all their time doing it, and that is simply unrealistic,” said Valentina Muraglie, who sits on the board of Italy’s association of large families.

Her own son, Antonio, put aside his collection of Harry Potter books and replaced reading with scrolling as a teenager.

Now in his 20s, he finds it hard to read in depth, which she blames on social media algorithms that sucked away his attention.

“Once he had a phone in his hand, at 16, little by little, books started to disappear,” she told Reuters. “Within a ​few years, he stopped reading altogether.”

The World Health Organisation warns that ​problematic social media use - marked by addiction-like behaviour - is increasing ⁠among adolescents and is linked to lower well-being, poor sleep and broader health risks.

Studies published in JAM Paediatrics, a US medical journal, point to measurable differences in brain development among heavy social media users, particularly teenagers whose brains are still developing.

The Italian case argues that social media platforms use reward mechanisms modelled on slot machines to foster dependency, by repeatedly triggering dopamine, a brain chemical linked to pleasure and reward.

“Each ‘like’ ​or notification triggers dopamine release, tying users to the platform in a way that resembles addiction,” said Tonino Cantelmi, a plaintiffs’ advisor and director of the School of Specialisation in Cognitive-Interpersonal Psychotherapy ​in Rome.

Families bringing the case say brain ⁠scan studies of social media users show activity in areas of the brain associated with addiction.

Asked about the scientific evidence on addiction presented in court, spokespeople from Meta and TikTok declined to comment on the litigation, while repeating their earlier comments on the companies’ records on mental health.

Some psychologists caution against drawing simple conclusions about the effects of social media on adolescents.

“The healthiest approach when dealing with adolescents is to accept that we are unprepared,” Federico Tonioni, head of the Web Psychopathology Centre at Rome’s Gemelli hospital, said.

He added he could not conclude that his patients would suffer less ⁠in a world ​without social networks, warning against over-reliance on parental control.

“If there is something dangerous, it is control over children. Young people need to be listened to. Control is ​not a healthy form of presence. The healthiest distance is trust.”

Irene Roggero Ugues said she joined the lawsuit to help ensure that other parents are made aware of risks that she did not learn about until after it was too late to save Rossella.

“We underestimated certain risks and didn’t know they existed, but others can still act. ​There’s no point keeping this to myself, and I don’t think Rossella would mind.”

Tuesday’s deal follows a blockbuster Nasdaq ​debut for the rockets-to-AI company last week, in which its valuation surged to more than $2 trillion.

The acquisition will give xAI, which was acquired ‌by SpaceX in February, a stronger hold in AI coding, one of the first areas where companies have turned AI into a real source of revenue from businesses.

Capitalizing on that interest is crucial for SpaceX as it had pitched its IPO investors an addressable market worth $28.5 trillion, the theoretical maximum revenue it could capture, of which a big share is expected to come from ​AI for businesses.

Cursor is one of several Silicon Valley startups that have drawn waves of developers by using AI to automate coding, making it ​a key rival to market leaders Anthropic and OpenAI. But a lack of access to computing power has hampered Cursor’s growth.

“Cursor ⁠does not have the scale of OpenAI or Anthropic, but it has built some very impressive coding models relative to cost. That makes this a positive move ​for SpaceX,” said Matt Britzman, senior equity analyst at Hargreaves Lansdown.

SpaceX had been eyeing Cursor for months and had in April unveiled an option to either buy ​the startup for $60 billion later this year or pay $10 billion for a partnership.

In its IPO filing, the company had said Cursor’s access to developers’ data, including coding requests and design decisions, could help improve its AI models such as Grok.

SpaceX said on Tuesday it would soon release an AI model on Cursor as well as Grok Build, xAI’s coding agent, which it has ​been jointly training for several months.

The all-stock transaction, for which SpaceX will not use proceeds from its IPO, is expected to close in the third quarter of ​2026.

Its shares jumped 10% in early trading, putting the company on track to add about $247 billion to its market capitalisation of $2.53 trillion. At $211.27, the stock has climbed more than 56% ‌from its ⁠IPO price of $135.

If the gains hold, SpaceX is set to overtake Amazon in market value to become the fifth-largest company.

HIGH-PROFILE BACKERS, RAPID GROWTH

Paying in stock, experts said, lets SpaceX take advantage of its towering valuation, which would mean giving up a relatively small slice of equity for a $60 billion deal.

“One of the things that makes SpaceX so valuable is how valuable it is. The Cursor acquisition costs materially less in dilution because of SpaceX’s high valuation,” billionaire Bill Ackman said in a post on ​X.

Cursor’s business has scaled rapidly since its ​founding in 2022, with roughly $2.6 billion ⁠in annualised business-to-business revenue and enterprise sales growing sharply, according to company data shared with Reuters earlier this month.

The San Francisco-based company, backed by prominent Silicon Valley venture capitalists such as Andreessen Horowitz and Thrive, as well as Nvidia and Alphabet’s ​Google, had reportedly been in talks for a funding round valuing it at $50 billion.

SpaceX will pay a termination fee of $10 ​billion if Tuesday’s deal ⁠collapses under specific circumstances. It will pay only $4 billion if the deal fails due to antitrust issues, according to the regulatory filing.

It was not immediately clear if the deal would affect SpaceX’s agreements to rent out its data centres.

The company has, in recent weeks, struck deals with Anthropic and Google to lease cloud computing capacity worth roughly $26 billion combined on ⁠an annual ​basis.

Both deals include 90-day termination clauses, meaning SpaceX could quickly reclaim computing capacity if needed.

“If usage ​of Grok and Cursor picks up enough, it can go back to using their capacity internally, but it appears that they will provide capacity to Anthropic and Google for the foreseeable future,” said D.A. ​Davidson analyst Gil Luria.

French consumer ​group UFC-Que Choisir sued Ubisoft in March after ​the video game maker shut down servers ⁠for its online racing game “The Crew”, making it ​permanently unplayable for buyers. The case is backed by ​the “Stop Killing Games” campaign launched after the controversy.

Ubisoft said players bought limited access, not full ownership. UFC-Que Choisir alleges ​the company misled consumers about how long the ​game would remain available and imposed unfair contract terms stripping ‌players ⁠of ownership rights.

The Commission said copyright and other intellectual property rules prevent it from imposing an obligation to keep games playable. It added it would ​work with ​consumer organisations ⁠and authorities to raise awareness of existing rights.

“Active enforcement of these existing consumer ​rights can also incentivise the providers ​to ⁠offer video games with longer lifespans and explore solutions for meeting consumer expectations,” the Commission said in ⁠a ​statement.

UFC-Que Choisir and Stop Killing ​Games did not immediately respond to requests for comment.

A cooling failure at a data centre near Chicago operated by CyrusOne — one of more than 55 facilities the company runs across the United States, Europe and Japan — had knocked out the CME Group, the world’s biggest exchange operator. Engineers were scrambled to get the cooling system back online.

The incident offered a rare, public glimpse into one of the defining infrastructure challenges of the AI era: how to stop data centres from overheating — and at what cost to the planet’s resources.

Why data centres run hot

The answer lies in the physics of computing. Data centres contain racks of servers stacked together, constantly switched on, consuming enormous quantities of power. AI and cloud servers crunching data generate intense heat that traditional air cooling systems are increasingly unable to handle.

“The chips that are in those data centres need to stay within certain temperatures, otherwise they either malfunction or they turn off,” said Daniel Mewton, a partner in the infrastructure, energy and natural resources practice at law firm Slaughter and May.

Cooling is not a minor operational expense — it is a defining one.

According to law firm White & Case, up to 40% of total energy consumption in data centres goes towards keeping them cool.

That has made the cooling sector a booming business in its own right.

In November 2025, power management firm Eaton announced it would acquire Boyd Corporation’s thermal business from Goldman Sachs Asset Management for $9.5 billion.

Competitor Vertiv has entered a $1 billion deal for PurgeRite Intermediate to expand its liquid cooling services.

The scale of the problem

The heat challenge is inseparable from the sheer scale of what is being built.

Last year, data centres consumed 448 terawatt-hours of electricity globally — more than the entire nation of Saudi Arabia uses in a year — with AI accounting for roughly a fifth of that total, according to a report from the United Nations University Institute for Water, Environment and Health.

The numbers are set to grow dramatically. Annual power consumption from data centres is projected to nearly double, reaching 945 TWh by 2030 — roughly equivalent to Japan’s entire national electricity consumption — with AI’s share rising to 40% of the total.

“The public debate still often treats AI as software, but AI is also physical infrastructure: data centres, electricity generation, cooling systems, transmission networks, chips, minerals, land and water,” said Kaveh Madani, the institute’s director and the report’s lead author.

The land footprint of data centres is also forecast to more than double, from around 6,900 square kilometres last year to over 14,500 square kilometres by 2030.

A thirst that cannot be ignored

Heat and electricity are only part of the burden. Cooling data centres requires vast quantities of water — and the AI boom is set to make that demand far more acute.

Data centres consumed 4.5 trillion litres of water last year, enough to meet the needs of more than 600 million people in Sub-Saharan Africa, the UN researchers found.

By 2030, that figure is expected to reach 9.3 trillion litres.

AI-driven growth alone could consume as much extra water each year as Americans currently drink, according to estimates from Ecolab, an $85 billion hygiene and water treatment specialist.

The standard method of cooling — evaporating water to remove heat — is the core of the problem.

When water evaporates, it effectively disappears from the local ecosystem.

AirTrunk, the Asia-focused Australian data centre operator acquired by Blackstone for $16 billion in 2024, acknowledged in its own sustainability report that 85% of the water it uses evaporates as it cools — a fact it initially described as water being “returned to the environment,” wording it later agreed to change after the misleading framing was pointed out.

The promise and limits of liquid cooling

The industry has begun responding. Liquid cooling, which can be up to 3,000 times more efficient than air at removing heat, is increasingly being adopted. More data centres are turning to water or specialised coolants circulated directly around server components, rather than cooling the surrounding air.

At the Consumer Electronics Show in January 2026, Nvidia chief executive Jensen Huang announced that the company’s new Vera Rubin chip system could be cooled using water at 45 degrees Celsius, compared to the current industry standard of around 6 degrees Celsius.

Because cooling water typically requires it to be chilled — a process that often involves evaporation — the higher temperature threshold significantly reduces water waste.

Microsoft has gone further, beginning construction of data centres that use “closed-loop” cooling systems, in which liquid circulates locally around the semiconductor housing rather than being used to cool ambient air.

The company says this approach can cut water consumption by 90%.

Microsoft has also financed a water treatment plant in Washington state allowing it to reuse data centre water.

Elon Musk’s xAI paid $80 million to treat and sell water contaminated by a former coal-fired power station near its Memphis AI data centre.

AirTrunk, meanwhile, taps recycled water for 55% of its intake and is building a wastewater treatment facility in Malaysia.

However, these advances address only part of the problem. The closed-loop systems being promoted by Nvidia and Microsoft apply to new facilities.

Some 80% of existing data centre sites handle older cloud-computing workloads that predate the AI boom and are unlikely ever to be upgraded to the latest cooling technology, according to S&P Global.

Cooling outages remain rare but not impossible. Data centres are contractually required to remain operational more than 99.99% of the time, making failures like the CyrusOne incident exceptional — but as the CME Group disruption demonstrated, even brief failures carry enormous consequences.

The bigger picture: water, energy and geography

On-site water savings also obscure a deeper problem. Data centres consume vast amounts of electricity, most of which still comes from gas and coal — power sources that are themselves water-intensive.

The Lawrence Berkeley National Laboratory estimated that US data centres directly consumed around 66 billion litres of water in 2023.

The indirect effect, through energy consumption, was 800 billion litres — more than twelve times larger.

Geography compounds the risk. Data centres tend to cluster together — in Arizona, Virginia and Singapore — and many of these hubs sit in areas already facing chronic water stress.

Almost half of all data centres are located in regions of high or very high water stress, per S&P Global, either because they are in arid zones or because available water has already been allocated to other users.

Phoenix, home to a major data centre cluster, is heavily dependent on the Colorado River, whose flow has been declining for 20 years and whose two main reservoirs are currently only 30% full.

A water utility in Melbourne, which endured a decade-long drought at the start of this century, is currently reviewing applications from 19 data centres that together are seeking permits for 20 billion litres of water a year.

Semiconductor fabrication plants — which produce the chips that power AI — add yet another layer of demand.

Taiwan’s TSMC is expanding in Arizona, where Intel also operates.

These so-called fabs require ultra-pure water to clean chips during manufacturing, recycle only a small percentage of what they use, and can consume as much water in a single week as a food processing plant uses in an entire year, according to one industry insider.

A race outpacing its solutions

The UN researchers are clear-eyed about the trajectory. “AI will not simply ‘run out’ of water or electricity worldwide,” Madani said. “

But in specific places, poorly planned data centre expansion could collide with existing resource pressures.

That is why responsible planning matters now, before infrastructure and dependencies become locked in.“

The broader warning is about governance as much as technology.

Unless governments respond to the rising environmental costs, the rapid rollout of AI infrastructure could strain land resources and generate mountains of electronic waste — on top of the power and water pressures already mounting.

“Right now, the competition for growing faster than others overshadows the very basic principles of sustainable growth,” Madani said.

The cooling sector’s deal-making boom, and the brief but telling panic that Huang’s CES announcement triggered — wiping $15 billion from the market value of five major heating, ventilation and air conditioning companies in a matter of days — suggest investors understand that the industry’s resource risks remain far from resolved.

All five companies have since recovered and trade well above their January lows.

Modine Manufacturing, once known for making tractor radiators, has risen 83%.

The message from the market is the same as the one from the UN: the heat problem is real, the solutions are partial, and the clock is running.

Prime Minister Sir Keir Starmer said keeping children off social media was the most effective way to protect them online.

The ban is expected to apply to major platforms including TikTok, Snapchat, Instagram, Facebook, YouTube and X, although the government has yet to publish a final list.

Officials said the restrictions would target platforms designed for social interaction and user-generated content.

Features such as livestreaming and contact with strangers will also face tighter controls.

Messaging services such as WhatsApp and Signal will not be included in the ban. YouTube Kids will also remain exempt.

The government is considering additional measures for users under 18, including overnight social media curfews and mandatory breaks to limit infinite scrolling.

Further details are expected next month.

AI-powered “romantic companion” chatbots will be restricted to users aged 18 and above, while other AI chatbot services will be required to limit intimate or adult-oriented features for minors.

The government plans to pass the necessary regulations before Christmas, allowing the restrictions to be implemented by spring 2027.

To enforce the rules, social media companies will be required to introduce robust age-verification systems, which could include facial age estimation technology or identity checks.

Media regulator Ofcom has been tasked with identifying the most effective verification methods.

The announcement follows a public consultation that received more than 116,000 responses.

The move has been welcomed by some bereaved families who have campaigned for stronger online protections for children.

However, critics argue that blanket bans could limit young people’s access to online communities and information.

Several social media companies have also opposed the proposal, warning that restrictions could push teenagers towards unregulated platforms with fewer safety measures.

The UK joins a growing number of countries considering or implementing age-based social media restrictions.

Australia introduced a similar ban in 2025, while countries including France, Canada, Norway, Denmark and Indonesia are pursuing comparable measures.

Australia became the first country to implement a nationwide ban, while several others are now pursuing similar measures.

The latest country to announce such restrictions is the United Kingdom, where Prime Minister Sir Keir Starmer has confirmed plans to ban social media use for children under 16 from 2027.

The proposed UK ban would apply to platforms including TikTok, Instagram, Facebook, Snapchat, YouTube and X, while messaging services such as WhatsApp and Signal would remain exempt.

The government is also considering additional safeguards, including overnight usage curfews and limits on endless scrolling for under-18s.

Australia led the way in December 2025 by introducing a ban on social media accounts for children under 16.

The restrictions cover major platforms including Facebook, Instagram, TikTok, Snapchat, X, YouTube, Reddit, Twitch and Threads, with companies facing significant fines for non-compliance.

Several other countries are at various stages of introducing similar legislation.

Austria is preparing to ban children under 14 from social media.

At the same time, Canada has proposed legislation that would prohibit social media use for those under 16 unless platforms can demonstrate sufficient child-safety protections.

Denmark plans to ban social media access for children under 15, with legislation expected to take effect as early as mid-2026.

France has also approved a bill restricting access for users under 15, although final legislative approval is still pending.

In Germany, policymakers are debating a proposal to prohibit social media use for children under 16, while Greece plans to introduce restrictions for users under 15 from January 2027.

Indonesia has announced plans to ban children under 16 from accessing major social media and online platforms, including TikTok, YouTube, Instagram and Facebook.

Malaysia is also preparing similar measures for users below 16.

Elsewhere, Poland, Slovenia, Spain and Turkey are either drafting or advancing legislation aimed at limiting children’s access to social networking platforms.

Supporters of the restrictions argue they are necessary to protect young people from harmful online content, excessive screen time and predatory behaviour.

Critics, however, have raised concerns about privacy, age-verification systems and the effectiveness of blanket bans, warning they may push children towards less regulated platforms.

Despite ongoing debate, governments across multiple regions continue to move forward with stricter controls on children’s access to social media.

The company received the export control directive to suspend access to Fable 5 and Mythos 5 for all foreign nationals, without being given specific details of its national security concern, Anthropic said in a statement.

It is Anthropic’s understanding that the government believes there is a method of bypassing, or “jailbreaking,” a safeguard that would prevent Fable 5 from being used in identifying software vulnerabilities, the company said.

The order comes just as a previous dispute between Trump administration officials and IPO-bound ​Anthropic showed signs of easing across parts of the US government.

Anthropic’s relationship with the government ruptured this year after it refused to allow the US military to use its ​AI models for domestic surveillance and fully autonomous weapons systems.

The government responded by putting Anthropic on a supply chain blacklist, set to take effect ⁠later in the year.

The action also marks a major escalation of US efforts to halt foreign adversaries’ AI capabilities.

For years, US export controls have focused on the chips and tools that power ​AI rather than on restricting foreign access to AI itself.

Anthropic said the government has given it only “verbal evidence of a potential narrow, non-universal jailbreak”.

“We disagree that the finding of a narrow potential jailbreak should ​be cause for recalling a commercial model deployed to hundreds of millions of people,” the company said.

The government directive and Anthropic’s response highlight growing tension between AI developers and regulators over how to assess risks from so-called “jailbreaks,” or methods used to bypass model safeguards.

As recently as Wednesday, Anthropic had called for greater US oversight of AI, including the ability to block models with unacceptable risks.

It said, however, that the government action on Friday did not follow ​principles of fair and fact-based regulation.

The Pentagon’s chief information officer, Kirsten Davies, said in a post on X that the Defence Department supported prioritising national security.

“Some things are simply more important than revenue ​cycles, clickbait, and pre-IPO valuation. America First. Always,” Davies said.

Anthropic confidentially filed for a US IPO last month, edging ahead of rival OpenAI in the race to reach public markets.

Sophisticated cyberattacks

Earlier this week, Anthropic rolled out an ‌AI model named ⁠Claude Fable 5, representing a new tier of capability it calls “Mythos-class.”

The model is accompanied by guardrails barring its use in risky areas such as cybersecurity, which some users have complained are “overly broad,” Anthropic said.

Experts have said that Mythos models, in the wrong hands, could dramatically accelerate sophisticated cyberattacks, particularly in sectors such as banking that rely on complex, interconnected, and often decades-old technology systems.

Anthropic said it had worked with the US government, among others, on safety ahead of the Fable launch and that models from rival AI providers showed a similar ability to unearth minor bugs in code.

“The net effect of ​this order is that we must abruptly disable ​Fable 5 and Mythos 5 for all ⁠our customers to ensure compliance. Access to all other Anthropic models will not be affected,” Anthropic said.

Anthropic said that it believed there was a “misunderstanding” and that it is working to restore access to the models as soon as possible.

“If this standard was applied across the industry, we believe it would ​essentially halt all new model deployments for all frontier model providers,” the company said.

Amazon’s cloud unit AWS said late on Friday that Anthropic has ​asked it to revoke access ⁠to the models for “all users in all regions.”

A US official confirmed that the Commerce Department had issued an export control directive to suspend all access to Fable 5 and Mythos 5 by foreign nationals.

Dean Ball, a former White House official who contributed to the AI Action Plan the administration issued in the summer of 2025, said in a post on X that the order suggests all “non-Americans” would be restricted from using ⁠Anthropic’s latest ​models, including those based in the US.

“This means you should expect to have to prove your citizenship to use Anthropic ​models,” Ball said.

Several key Anthropic personnel, including co-founder Chris Olah, AI researcher Andrej Karpathy and philosopher Amanda Askell, were born outside the United States.

Reuters was unable to determine their citizenship status, and an Anthropic spokesperson declined to comment on whether ​such staff would lose AI model access.