The intermediary buyer was an unnamed Southeast Asian firm referred to by prosecutors as “Company-1,” which Bloomberg identified as ​Bangkok-based OBON Corp, citing sources.

Alibaba Group Holding was among the end customers, the report ​added.

In an emailed statement to Reuters, an Nvidia spokesperson said the ⁠company expects its ecosystem partners to adhere to strict compliance at every ​level, stating that it will continue working with the government to enforce the rules.

Separately, ​Alibaba told Reuters it has no business ties with Super Micro, OBON or any third-party brokers cited in the indictment, and said banned Nvidia chips have never been used in its ​data centres.

Super Micro did not immediately respond to Reuters’ requests for comment, while ​OBON could not be immediately reached.

In March, the US Justice Department charged Super Micro co-founder ‌Yih-Shyan ⁠Liaw, sales manager Ruei-Tsang Chang and contractor Ting-Wei Sun with running a scheme to route US-made servers through Taiwan to Southeast Asia, where they were repackaged into unmarked boxes and smuggled into China.

Prosecutors alleged at least $2.5 billion in US AI technology was ​moved, including more ​than $500 million shipped ⁠between April and mid-May 2025.

Some of the $2.5 billion servers sold to OBON allegedly went to Alibaba, the report added.

In 2022, ​the United States banned the export of high-end chips from Nvidia ​to China ⁠amid concerns that they could be used for military purposes, but approved sales of Nvidia’s second-most powerful H200 chips in January this year under certain conditions.

Separately, Super Micro shareholders sued ⁠the ​Silicon Valley server maker in March, accusing it ​of securities fraud by allegedly concealing its reliance on sales to China that violated US export laws.

The Swiss-US maker of keyboards, mice, and video-conferencing equipment is betting on gaming, business customers and artificial intelligence-enabled devices to maintain growth, after cutting costs last year to offset the impact of US ​President Donald Trump’s tariffs.

The push comes despite supply disruptions in the Middle East, which have complicated ​shipments and are expected to cost the company about $15 million in sales in the ⁠current quarter, after a $5 million hit in the three months to the end of March.

Logitech expects momentum ​from its fourth quarter to carry into the current period, aiming for 2% to 4% sales growth in constant ​currencies to $1.190 billion to $1.215 billion.

“We can, and we should invest,” Faber told Reuters. “The world is changing so fast with AI, which offers so many opportunities.

“We came out of the last fiscal year with such a strong financial base, so we have the ​firepower to do it,” she added.

Logitech plans to keep total operating expenses for the fiscal year toward the ​top end of its long-term range of 24% to 26% of sales, up from 24.8% in the 12 months to ‌March ⁠2026.

Gaming and business demand are seen as resilient

Research and development spending to produce new devices should be around 6% of sales this year, after coming in slightly below that level last year, while sales and marketing spending will also rise from about 16%, Faber said.

Gaming remains a key focus, with younger consumers spending more time playing ​computer games, making it a ​resilient market, she added.

Logitech ⁠is also stepping up efforts to win more business customers, with demand expected to remain strong as companies boosted by strong recent earnings invest in new computer ​hardware.

The company would look at healthcare, education and government as long-term growth areas, ​Faber said.

Logitech ⁠has been shielded from oil price rises, which have made plastic more expensive, due to 78% of the company’s products using recycled rather than virgin plastic.

Still, Middle East disruption has hit sales because some products could not be ⁠delivered from ​factories in Asia to its distribution centre in Dubai and ​on to other parts of the Gulf and Africa, despite demand remaining intact.

“We’re not seeing that demand for our products is down,” Faber ​said. “It’s just logistically hard to get it to people.”

Co-founded by former Mistral researcher Theophile Gervet, the company said its GENE-26.5 model can run a range of robots, including those made by other companies. It is ​in advanced talks with potential customers in France, Germany and Italy.

The launch - alongside the robotic ​hand capable of tasks such as chopping tomatoes and solving a Rubik’s Cube - ⁠comes as Europe pushes to reindustrialise and cut reliance on Asian manufacturing.

Demand for industrial robotics is ​also rising. Germany’s Schaeffler said this week it expects its robotics order book to reach hundreds of millions ​of euros by 2030.

Founded in early 2025, Genesis AI raised $105 million in an initial funding round, one of France’s largest and matching the record seed round of Mistral AI - Europe’s leading AI company. Backers also include state investment bank ​Bpifrance.

Europe focus

Gervet told Reuters the company was prioritising Europe.

“There were two big reasons. The first one ​was the talent base,” he said. “The second reason was the industrial base as a market for us.”

Genesis is targeting ‌sectors such ⁠as automotive, electronics, pharmaceuticals and logistics, where conventional robots struggle with delicate or variable tasks such as wire harnessing, which involves bundling and taping cables.

The company said it is signing customers but declined to name them. Engagements will typically run three to five years, depending on client needs, said Vivian ​Sun, vice president of commercial ​and strategy.

It is ⁠also working with partners to build robotics datasets, including collecting real-world data from tens of thousands of industrial workers using sensor-equipped gloves.

Human-like hand

Genesis’ robotic hand ​is designed to more closely mirror human anatomy than standard grippers, enabling more ​direct transfer ⁠of human motion to machines.

In a video seen by Reuters, the robot cut tomatoes, cracked eggs, solved a Rubik’s Cube and played the piano.

The launch puts Genesis in competition with China’s Linkerbot, which Reuters reported is ⁠targeting a $6 ​billion valuation as demand grows for highly dexterous robotic hands.

Both ​companies are developing hardware to enable more human-like manipulation in industrial settings.

Genesis said it expects to raise more capital, but that ​a public listing remains premature.

The chatbot, which is available as a standalone app, website, and through X, recorded a sharp fall in usage and has fewer users than at the start of 2026.

In January, Grok ranked second globally among AI chatbot apps in terms of average daily active users, behind ChatGPT.

However, by April, it had fallen to fifth place, overtaken by Claude, Gemini and DeepSeek.

On mobile, Grok’s average daily active users fell from 13.9 million in March to 12.2 million in April, a 12.5% month-on-month decline.

In the United States, usage dropped from 1.4 million to 1.1 million over the same period, a 15.6% fall.

On the web, Grok also declined, with global daily visits falling from 10.5 million in March to 9.3 million in April.

US web traffic similarly dropped from 2.3 million to 2.1 million.

By contrast, rival platforms saw strong growth. Claude’s global mobile users rose from 16 million to 23 million, while Gemini increased from 12.4 million to 14.8 million.

On the web, Claude surged from 19.8 million to 27.5 million daily visits, while Gemini rose from 83.7 million to 92.1 million.

Despite the short-term decline, Grok’s overall usage remains higher year-on-year, with web visits up 42.4% and mobile users up 83.5% compared to April 2025. However, competitors have posted far stronger growth rates.

ChatGPT continues to dominate the AI chatbot market.

In April, it recorded 244.9 million average daily mobile users, far ahead of Claude, Gemini, DeepSeek and Grok.

It also led web traffic with 183.7 million daily visits, roughly double Gemini’s figures.

The fluctuations in Grok’s user base come amid reports of product changes, subscription limits, and internal challenges at xAI, which is led by billionaire Elon Musk.

Separately, Musk has been involved in an ongoing legal dispute with OpenAI CEO Sam Altman over the company’s structure and founding mission, further highlighting tensions among major players in the AI industry.

The testimony ‌by Greg Brockman came in the second week of a trial in California that could determine the future of OpenAI, which sparked a widespread craze over generative artificial intelligence after launching its ChatGPT chatbot in late 2022.

OpenAI plans to spend $50 billion on computing resources in 2026, Brockman said in court.

Musk accused OpenAI and Chief Executive Sam Altman of conning him into giving $38 ​million to the nonprofit, only to see it abandon its charitable goals and become a for-profit company to enrich themselves.

The world’s richest person is ​seeking $150 billion in damages to be paid to the nonprofit, and for Altman and Brockman to be removed from their ⁠leadership roles. Musk left OpenAI’s board in February 2018.

Musk wanted full control

In his second day of testimony, Brockman said that in 2017, Musk had wanted ​OpenAI to change its corporate structure because it was too hard for a nonprofit to raise the amount of money OpenAI required to build advanced AI models.

Brockman ​said the Tesla and SpaceX founder made clear that he wanted to become OpenAI’s leader if that happened.

Altman was the only other candidate, Brockman said.

Brockman described a particularly intense meeting in which Musk said he deserved a majority stake in OpenAI because of his business experience.

Musk said he intended to use that stake to build a self-sustaining city on Mars, according ​to Brockman.

“He said he needed $80 billion to create a city” on Mars, Brockman said.

“In the end, he needed full control.” Brockman added that Musk said he ​would decide when to relinquish full control.

Brockman said the meeting with Musk in August 2017 started well.

He said Musk had recently given Teslas to some OpenAI employees in gratitude ‌for their ⁠work, and former OpenAI chief scientist Ilya Sutskever painted a portrait of a Tesla to give to Musk as a token of thanks.

But according to Brockman, Musk grew angry when discussing a potential equity structure for OpenAI that he didn’t like, saying, “I decline.”

Brockman said Musk stood up and walked over so fast he was concerned Musk would hit him, but instead Musk picked up Sutskever’s painting and stormed out, saying he would withhold new funding until matters were sorted.

Incentive to go to Mars

Musk’s lawyers have tried ​to portray Brockman as also seeing ⁠dollar signs when looking at OpenAI.

On Monday, Brockman testified that his stake in OpenAI is worth almost $30 billion.

He also said he holds stakes in two startups backed by Altman, and a 1% stake in Altman’s family fund.

Evidence in the case also includes ​a 2017 diary entry where Brockman wrote: “Financially, what will take me to $1B?”

In March 2019, OpenAI restructured as a for-profit ​unit governed by the ⁠nonprofit, allowing it to accept money from outside investors.

The business has since raised more than $100 billion to hire researchers, buy computing power and expand ahead of a potential $1 trillion initial public offering this year.

OpenAI has said Musk was embittered because he left its board before the company’s successes, and now wants control.

It also said Musk sued ⁠to bolster ​his own AI company xAI, which merged with SpaceX in February.

SpaceX may also go public this year, ​in an IPO that could be larger than OpenAI’s.

According to a registration statement, SpaceX’s board in January approved awarding Musk 200 million of super-voting restricted shares if its market value reaches $7.5 trillion and it ​creates a permanent colony on Mars with at least 1 million people.

Justice Elena Kagan, on behalf of the court, declined to pause a ruling by the San Francisco-based 9th US Circuit Court of Appeals that deemed Apple in contempt in the Epic lawsuit contesting ​App Store fees. Apple had sought the delay to give it time to file a full ​Supreme Court appeal of the 9th Circuit decision.

Apple and Epic have clashed for years over ⁠the rules governing Apple’s App Store. The contempt ruling and the scope of Apple’s court-ordered obligations are ​the latest issues in the dispute to reach the Supreme Court. Apple has said the 9th Circuit decision would ​affect how millions of app purchases are made.

Epic Games won the contempt order last year as part of litigation it brought in 2020 seeking to loosen Apple’s control over transactions in applications that use the company’s iOS operating system and its restrictions ​on how apps are distributed to consumers.

Apple mostly defeated Epic’s lawsuit, but was required in a 2021 court ​injunction to let developers include links in their apps directing users to non-Apple payment methods.

Apple allowed the links but adopted ‌new restrictions, ⁠including a 27% commission on developers for purchases made on payment systems outside the App Store within seven days of clicking a link. Apple charges developers a 30% commission for purchases within the App Store.

Epic argued that the new 27% commission flouted the earlier injunction. In 2025, US District Judge Yvonne Gonzalez Rogers found Apple in civil contempt for violating the ​injunction.

The 9th Circuit in ⁠December upheld the judge’s contempt finding but allowed Apple to make new arguments about what commission it should be allowed to charge for digital goods bought in apps ​distributed through the App Store but paid for using third-party payment systems.

Apple has denied ​violating the ⁠judge’s order and has argued that the injunction should not be applied to millions of developers beyond Epic Games.

“Regulators around the world are watching this case to determine what commission rate Apple may charge on covered purchases in huge ⁠markets ​outside the United States,” Apple told the Supreme Court in a filing.

Epic ​has argued that Apple should not be allowed to sidestep the judge’s original injunction, saying this would “give Apple more time to continue unfairly ​profiting at the expense of consumers and app developers.”

The lawsuit, filed by Peter Landsheft in US federal court in California in 2024, arose after the iPhone ​maker announced — and started running advertisements for — a bevvy ​of AI upgrades at its annual software ⁠developer conference in 2024, saying they would become ​available with new iPhones that fall.

The iPhones launched without ​those features, which the plaintiffs claimed harmed shareholders.

In 2025, Apple said that the AI overhaul of Siri would not come ​until this year, and executives have now confirmed ​that the new Siri features will be unveiled at Apple’s annual ‌developer ⁠conference next month.

Apple did not admit to any fault in the settlement, which still needs approval from a judge.

In a statement, Apple said it ​released numerous other ​AI features ⁠since the launch of what it calls Apple Intelligence in 2024.

“Apple has reached ​a settlement to resolve claims related to ​the ⁠availability of two additional features. We resolved this matter to stay focused on doing what we do ⁠best, delivering ​the most innovative products and ​services to our users,” the company said in a statement.

The Centre for AI Standards and Innovation at the Department of Commerce said on Tuesday that the agreement would allow it to evaluate the models before deployment and conduct research ​to assess their capabilities and security risks.

The agreement underscores growing concern in Washington ​over the national security risks posed by powerful AI systems. By ⁠securing early access to frontier models, US officials are aiming to identify threats ranging ​from cyberattacks to military misuse before the tools are widely deployed.

The development of advanced AI ​systems, including Anthropic’s Mythos, has in recent weeks created a stir globally, including among US officials and corporate America, over their ability to supercharge hackers.

“Independent, rigorous measurement science is essential to understanding frontier AI ​and its national security implications,” CAISI Director Chris Fall said in a statement.

The move ​builds on agreements with OpenAI and Anthropic, established in 2024 under the Biden administration when CAISI ‌was known ⁠as the US Artificial Intelligence Safety Institute.

CAISI, which serves as the government’s main hub for AI model testing, said it had already completed more than 40 evaluations, including on cutting-edge models not yet available to the public.

Developers frequently hand over versions of their ​models with safety guardrails ​stripped back so ⁠the centre can probe for national security risks, the agency said.

Microsoft and xAI did not immediately respond to requests for comment. Google ​declined to comment.

Last week, the Pentagon said it had reached agreements ​with seven ⁠AI companies to deploy their advanced capabilities on the Defence Department’s classified networks as it seeks to broaden the range of AI providers working across the military.

The Pentagon announcement did ⁠not ​include Anthropic, which has been embroiled in a dispute with the Pentagon over guardrails on the military’s use of its AI tools.

Tech companies ​are coming under increasing pressure worldwide to ⁠come up with age-checking measures over mounting concerns about ​online abuse, teen mental health and the spread of ​AI-generated child sexual images.

Meta last year rolled out technology to proactively find accounts they suspect to be teens, even if ​they list an adult birthday, and place them ​in Teen Account protections.

“This technology will be expanded to 27 countries ‌in ⁠the European Union. Meta is also expanding this technology to Facebook in the United States for the first time, with the UK and EU to ​follow in June,” ​the company ⁠said in a blogpost.

It also detailed its use of advanced artificial intelligence to ​detect underage accounts beyond simple admissions of ​age.

This ⁠includes using AI technology to analyze entire profiles for contextual clues to determine if an account likely ⁠belongs to ​someone underage and strengthening circumvention ​measures to prevent new accounts from users Meta suspects are underage.

Beijing-based Linkerbot completed what it called a “series B+” funding ‌round last week that valued it at $3 billion.

It did not say when the next funding round will be launched or whether it was targeting the previously undisclosed $6 billion valuation in a private investment round or in an initial public offering.

Prominent early backers of the two-year-old unicorn include Alibaba’s Ant Group and Sequoia spin-off HongShan Group, while the state-backed Zhongguancun Science Park Fund, Bank ​of China Asset Management and Fosun Capital participated in the latest round, the company said in a statement on Thursday.

Linkerbot currently holds over 80% ​of the global market share in high-degree-of-freedom (DoF) robotic hands, and plans to scale production “soon” to 10,000 units a month from almost ⁠5,000 currently, CEO Alex Zhou told Reuters in an interview.

Investor interest in China’s humanoid robotics industry has surged this year after market leaders such as Unitree demonstrated ​their products’ staggering technical advances during a widely watched TV performance and the Beijing humanoid robot half-marathon last month.

Unitree filed for a Shanghai IPO in March, seeking a valuation of ​up to $7 billion.

Unlike competing humanoid makers such as X Square Robot that focus on training robotic hands for household chores, Linkerbot specialises in high-value human craftsmanship.

“We aren’t just making hands. Our goal is to replicate the entire library of human dexterous skills within our hardware,” Zhou said, referring to the firm’s LinkerSkillNet platform, which he claims is the world’s largest real-world dexterous manipulation dataset.

The platform ​is a multimodal data collection system that converts human skills into standardised, reusable capabilities for robotic hands, containing over 500 skills so far.

“The hand is the most ​complex part of the whole humanoid robot. Elon Musk described on several occasions that the part was taking more than half of their whole engineering effort for Tesla’s Optimus,” said Georg ‌Stieler, head ⁠of robotics and automation at technology consultancy Stieler.

Musk promised that the latest Optimus model, to be launched this spring, will have “the manual dexterity of a human”.

Technical frontier

Inspired by his childhood fondness for Doraemon, the Japanese cartoon robotic cat who holds an infinite array of gadgets in his pocket, CEO Zhou envisions his robots playing the piano, giving massages or even doing dentistry: skills he says are a “value-add that is at least triple that of basic labour”.

Linkerbot’s hands can already rapidly turn screws, grasp deformable ​soft objects, thread a needle and engage ​in high-precision manufacturing.

The company supplies ⁠some of China’s leading humanoid robot makers as well as some foreign industrial giants, which the company declined to disclose due to NDAs.

Its basic O6 light-weight model can carry a 50 kg load despite weighing only 370 g, performance Zhou said ​was a key advantage for industrial applications where miniaturisation and strength are required.

The company manufactures key components like joint modules, ​motors and reducers in-house, ⁠and uses specialised polymers that are self-lubricating and corrosion-resistant, Zhou said.

Besides industrial settings, Linkerbot’s hands are used by research institutes and leading global universities.

The company has over 400 employees and five factories in Beijing and Shenzhen, and is developing intelligent production lines where robotic hands manufacture other hands.

A major obstacle to the widespread factory use of humanoids is ⁠their cost, ​at $100,000 to $150,000 per unit for the leading industrial models from Unitree, AgiBot and UBTech, analysts say. But ​Linkerbot says its hands are more easily deployed.

“Chinese factory owners are extremely pragmatic. They’ve realised that for most factory work, two arms and a pair of dexterous hands are enough,” said Zhou.

“Currently, many of ​our customers simply mount our hands onto existing robotic arms rather than buying a full humanoid,” he said.

Eight new AI partnerships were ​signed in the past six months alone, according to research firm Ampere Analysis.

Among them, the nine-time constructors’ champion Atlassian Williams F1 team are partnering with ‌AI company Anthropic for its Claude model to support team operations and race strategy.

“It’s much more than a sticker on a car or a sticker in a billboard,” Williams’ Board Advisor Peter Kenyon told Reuters.

“We see it as one of our differentiating points: how can this partner help us in that journey back to the top?”

Whereas F1 cars in yesteryear had a plethora of brands with ​tobacco companies at the centre, now partnerships often centre on AI and tech companies helping the teams understand datasets, while benefiting from great exposure.

“What Anthropic ​and our tech team are doing is understanding the opportunities and then integrating those into our business to be able to ⁠demonstrate for ourselves and them, and showcase their technology in the pursuit of getting Williams back to the top,” Kenyon added.

AI can be a key tool enabling teams ​to navigate new regulations and new cost cap rules, now set at $215 million.

“Efficiency is one of the ubiquitous benefits of AI products, meaning a natural synergy between teams and AI ​brands,” said Adam Lewis, a senior analyst from Ampere Analysis.

Technology led the top 10 spending categories for F1 teams, reaching an estimated $769 million last season, up 41% from the previous year, according to intelligence platform SponsorUnited.

AI and machine learning brands account for four of the top 15 new sponsorship investors, a SponsorUnited report also showed, including a $65 billion-valued cloud infrastructure company, CoreWeave, which has a partnership ​with the Aston Martin F1 team.

In the 2025 season, the single-seater motorsport reached $2.54 billion in total team sponsorship and was the second-highest-grossing sports property behind America’s National ​Football League, which achieved $2.7 billion.

Helping with rules

AI has been innovative in sifting through administrative tasks and interpreting key rules within sporting and technical regulations, helping engineers make ‌swifter decisions ⁠during on-track situations which were impossible decades ago.

“So it’s gone from a sort of basic AI to more of an agentic approach where rather than just searching for something, it’s actually providing decisions for us,“ Jack Harington, the group partnership lead for Oracle Red Bull Racing, told Reuters.

The Red Bull outfit, which four-time champion Max Verstappen races for, has a partnership with $494 billion-valued software company Oracle and has embedded its technological nous across the team.

“So it’s really playing into the strength of AI as an enabler for our ​team. Allowing them (engineers) to focus on the ​core responsibilities they have and perform ⁠better at what they do,” Harington added.

Technology companies like Alphabet-owned Google are also seeing positives from entering the F1 arena.

“These blue-chip companies are using Formula One as a launchpad and spotlight for their own AI products or rebrandings,” Lewis said, noting Google’s partnership ​with F1’s McLaren shifted to Google Gemini, a generative AI tool, from Google Pixel.

As an organisation, F1, which returned to Miami ​after no races in ⁠April, has also embraced AI. Its partnership with Amazon Web Services uses generative AI for live television broadcasting, and in 2024, it applied generative AI to the design of the Montreal trophy after it was crafted by a silversmith in the United Kingdom.

“I think F1 has the never-ending, unquenchable thirst for the latest technology,” Lenovo’s Global Chief Information Officer Arthur Hu told Reuters.

Lenovo, ⁠a Hong ​Kong-listed technology company, is one of F1’s global partners and has been in a partnership with the ​organisation since 2022.

Hu said that Lenovo helps F1 to enhance productivity, mobility and remote collaboration through Lenovo laptops and devices, including AI PCs, to support the delivery of races.

“Formula One is at the sweet ​spot where it’s an intensely technical sport … And so I think that only opens up new possibilities,” Hu said.

Among the most important updates is the ban on WhatsApp use in banking services that came into effect from May 1.

Under new guidelines, the UAE Central Bank has prohibited financial institutions from using WhatsApp for customer services.

The restrictions cover sharing sensitive financial information, confirming transactions, and sending one-time passwords (OTPs).

The goal of the ban is to protect users from fraud and ensure that financial institutions comply with data privacy laws.

Under the new instructions, banks and payment providers are required to use secure communication channels like mobile apps and call centres.

In a bid to ensure national security and protect personal freedom, the authorities said that private chats are not exempt from the country’s strict cybercrime laws.

Forwarding unverified content, sharing images without consent, or tagging people in defamatory posts can lead to fines of up to Dh500,000 or imprisonment, they warned.

Legal experts emphasise that forwarding a message is considered ‘re-publication’, making users liable under the law.

In a similar vein, the authorities warned that WhatsApp group administrators could be held responsible for any illegal/offensive content shared within their groups.

Under the UAE’s laws, WhatsApp group administrators are responsible for removing illegal messages and removing members who post harmful content.

Any failure to do so could lead to legal consequences, experts warned and added that it’s crucial for admins to respond promptly when they come across prohibited material in their groups.

Legal experts said that WhatsApp messages can be used as court evidence.

They said that in a recent ruling, Dubai’s Court of Cassation stated that WhatsApp messages could be presented in courts as evidence in legal disputes.

Legal experts warned that following the court instructions, users should be cautious with their messages, as they may be produced in court, especially in cases involving personal disputes or financial matters.

The case, which will be tried before a judge in Santa Fe, stems from a lawsuit filed by New Mexico Attorney ‌General Raúl Torrez, a Democrat, accusing the social media giant of designing its products to addict young users and failing to protect children from sexual exploitation on its platforms.

At the heart of the trial is whether Meta’s platforms have created a “public nuisance” under New Mexico law.

That finding would allow the judge to order wide‑ranging remedies aimed at curbing alleged harms to young users.

The case is being closely watched as states, municipalities and school districts across the country ​pursue similar claims seeking to force changes at the industry level.

Monday’s trial marks the second phase of New Mexico’s lawsuit.

A jury in March found Meta violated the state’s consumer protection ​law by misrepresenting the safety of Facebook and Instagram for young users.

It ordered the company to pay $375 million in damages.

Criticism of children’s safety on ⁠social media has been mounting for years.

On Wednesday, Meta warned investors that legal and regulatory blowback in the European Union and the US “could significantly impact our business and financial results.”

Sweeping remedies at stake

Torrez’s ​office is expected to seek both billions of dollars more in damages and an order requiring Meta to make substantial changes to its platforms for New Mexico users, according to court filings.

Meta has said it ​has already addressed many of the state’s concerns and taken extensive measures to ensure its young users are safe.

The company said in court filings last week that many of the changes Torrez’s office is seeking are impossible for it to comply with and may force it to withdraw from the state entirely.

“The New Mexico Attorney General’s focus on a single platform is a misguided strategy that ignores the hundreds of other apps teens use daily,” a ​Meta spokesperson said in a statement ahead of the trial.

“Rather than providing comprehensive protections, the state’s proposed mandates infringe on parental rights and stifle free expression for all New Mexicans.”

A ‘Public nuisance’

The trial before ​Judge Bryan Biedscheid will examine whether Meta’s conduct meets the standard for a public nuisance under New Mexico law, which would allow the court to impose remedies aimed at abating the alleged harm.

A public nuisance claim targets ‌activities that unreasonably ⁠interfere with the health and safety of a community.

Classic examples include blocking a public road, polluting a waterway or emitting noxious fumes.

State governments have invoked public nuisance law in recent decades to pursue a broader range of industries, including litigation tied to tobacco, opioids, climate change, and vaping, said Adam Zimmerman, a professor at USC’s Gould School of Law.

New Mexico’s case is among a growing number of lawsuits accusing Meta and other social media companies of intentionally designing products to be addictive to young people.

While many cases have been filed by families over specific injuries to individuals, more than 40 other ​states and over 1,300 school districts have filed ​lawsuits seeking court-ordered changes and damages under public ⁠nuisance law.

New Mexico said it plans to ask the judge to order Meta to make changes, including verifying users’ ages, redesigning its algorithm to promote quality content for minors, and ending autoplay and infinite scrolling for minors.

“It will be an opportunity for us to explore more deeply the size and scale and ​effectively the monetary value of the public nuisance harm that was a product of this business’s behaviour for the last, you know, 10 ​or 15 years,” Torrez told ⁠reporters at a press conference on Thursday ahead of the trial.

The company has said in court filings that it cannot have created a public nuisance because it has not interfered with a public right.

It also said there is no scientific evidence to support the idea that social media has caused mental health problems, and that many of the state’s requests are “technologically impractical or completely impossible.”

In a public nuisance case, the state ⁠can also seek ​money damages to abate the harm.

That sum could be substantial when the impact is said to have affected large segments ​of the population.

Torrez’s office has not detailed the amount it will seek.

Meta said in court filings that New Mexico plans to ask for $3.7 billion in damages to fund a 15-year mental health plan, including new healthcare facilities and hiring providers, a request ​it said would require it to pay for mental health care for all teens in the state, regardless of the cause of their needs.

Musk, the CEO of Tesla and SpaceX as well as the world’s richest person, is also ​suing OpenAI’s Chief Executive Sam Altman and its President Greg Brockman, saying they betrayed him and the public by abandoning the mission to be ‌a benevolent steward of AI for humanity.

Musk recasts OpenAI as a ‘charity’

The word “charity” doesn’t appear once in the 2015 announcement of the formation of OpenAI as a nonprofit artificial intelligence research company. But Musk repeatedly described OpenAI as a charity and testified that Altman and Brockman reneged on an initial promise to keep the nonprofit model.

“It was specifically meant to ​be for a charity that does not benefit any individual person. I could’ve started it as a for-profit, and I specifically chose not to,” Musk testified.

Musk says OpenAI wouldn’t exist without him

Building an AI research lab requires top-tier talent and considerable computing power. Musk said OpenAI relied on ⁠his connections for both.

“I came up with the idea, the name, recruited the key people, taught them everything I know, provided all of the initial funding,” Musk said.

Musk ​said he recruited Ilya Sutskever, a top researcher, from Google while that company’s founders, Larry Page and Sergei Brin, repeatedly tried to entice him to stay.

“After I recruited Ilya to ​OpenAI, Larry Page refused to speak to me ever again,” Musk testified.

On computing power, Musk said that OpenAI relied on his connections with Microsoft CEO Satya Nadella and Nvidia CEO Jensen Huang. “The only one who could actually call Satya Nadella and have him pick up was me,” Musk said. “The only reason he’s in this thing is because of me. Those are his words.”

Musk spoke about AI safety

Musk ​testified that he learned from discussions with Larry Page that the Google founder lacked concern about AI safety.

“I said, ‘What if AI wipes out all humans?’ He said that would ​be fine so long as artificial intelligence survives. I said that was insane, that’s just crazy. And then he called me a ‘speciesist’ because I care about humans more than AI. … The reason ‌OpenAI exists ⁠is that Larry Page called me a ‘speciesist.’ … What would be the opposite of Google? An open-source nonprofit.”

‘It felt like a bride’

Musk said he asked Altman in late 2022 about an investment of $10 billion in OpenAI by Microsoft, which Musk described as a “bait and switch” in a text message shown to jurors.

Altman responded, “I agree this feels bad.” Altman then offered Musk an opportunity to buy stock in OpenAI, which Musk said, “frankly, it felt like a bribe.”

Musk, on training his own AI company

Musk was asked why he used OpenAI to train ​his xAI company if he considered OpenAI’s model ​a danger. “It is standard practice to ⁠use other AIs to validate your AI,” Musk said. Asked why his company was not structured as a charity, Musk testified, “For-profit companies can be socially beneficial.”

‘We all could die’

Musk’s cross-examination by William Savitt, a lawyer for the OpenAI defendants, was tense at times. ​Musk accused Savitt of asking tricky and leading questions, which the judge said was permitted. Savitt was admonished for not allowing ​Musk to finish his ⁠thoughts. “Few answers are going to be complete, especially when you cut me off all the time,” Musk said.

There was also pre-trial tension when Musk’s lawyers wanted to be able to question an expert witness about the extinction risk of AI, something OpenAI opposed. “Extinction risk is a real problem. This is a real risk. We all could die,” said Musk’s lawyer, Steven Molo.

The judge ⁠limited the scope ​of the expert’s testimony and said that she thought “it’s ironic that your client, despite these risks, is ​creating a company that’s in the exact space.”

The report, published by Israeli daily Yedioth Ahronoth, describes an incident over the Lebanese town of Taybeh in which Israeli defence systems were reportedly unable to detect or disrupt incoming unmanned aerial vehicles (UAVs).

According to the report, an Israeli medical evacuation helicopter carrying wounded soldiers came under threat from a drone during a rescue operation.

Ground troops reportedly opened fire at the approaching UAV, which detonated close to the aircraft.

The drones are said to be first-person view (FPV) systems operated via fibre-optic cables rather than radio signals, making them resistant to electronic warfare and jamming technologies.

The cable reportedly extends for tens of kilometres, allowing operators to guide the drones directly to targets.

Military analysts cited in the report said the systems are also difficult to detect due to their lightweight fibreglass construction, which reduces radar and thermal signatures.

The report claims the drones have been used to strike armoured vehicles, including Israeli tanks equipped with the “Trophy” active protection system, which is designed to intercept incoming threats.

One reported attack in Taybeh involved an explosive drone striking an armoured unit, killing one soldier and injuring several others.

A subsequent drone reportedly detonated near a helicopter during evacuation efforts.

Israeli field commanders quoted in the report described limited countermeasures, relying largely on visual detection and small-arms fire.

Some units have reportedly begun using improvised protective measures, including netting over positions and vehicles.

Military sources acknowledged that existing defences were not fully prepared for this type of threat, despite similar drone warfare tactics being observed in other recent conflicts.

Analysts also noted that while fibre-optic drones offer improved resistance to jamming and precise control, they remain vulnerable to weather conditions and physical obstructions that can sever the cable link.

The development reflects an ongoing shift in regional warfare, where low-cost drone systems are increasingly being used alongside traditional weapons to challenge more advanced military platforms.

In testimony on Tuesday before a nine-person jury in Oakland, California, federal court, the world’s richest person sharply criticised the 2019 decision by the nonprofit OpenAI co-founder and Chief Executive, Sam Altman and its ​President, Greg Brockman, to create a for-profit entity.

“If we make it okay to loot a charity, the entire foundation ​of charitable giving in America will be destroyed,” Musk testified.

OpenAI has said it created a for-profit ⁠entity to allow it to buy computing power and pay top scientists.

Its lawyers have argued that Musk is motivated by ​a compulsion to control OpenAI and bolster his own AI company, SpaceX unit xAI, which lags OpenAI in user adoption.

Judge scolds Musk over X posts

The trial highlights the depth of the rupture between Musk and Altman. The two Silicon Valley icons once partnered in the quest to develop the fast-growing AI technology, a pillar of growth in the US economy that is also fuelling anxiety about job losses.

The pair co-founded OpenAI in ​2015 to create a benevolent steward of the technology and fend off rivals such as Alphabet Inc’s Google. Musk, the chief ​executive of Tesla and SpaceX, left OpenAI in 2018 after investing $38 million. Microsoft, also a defendant, invested $10 billion in OpenAI in 2023.

On Wednesday, Musk, ‌54, will ⁠resume being questioned by his own lawyer.

He is then expected to be cross-examined by lawyers for OpenAI and the other defendants, who have argued that AI safety was not a priority for Musk when he was with the company and that he derided employees who focused on it as “jackasses.”

Before jurors were seated on Tuesday, US District Judge Yvonne Gonzalez Rogers admonished Musk after OpenAI ​lawyers complained about posts on ​X in which Musk assailed ⁠Altman as “Scam Altman.” Musk, known for brash public commentary, agreed to minimise his social media activity, as did Altman.

Musk seeks $150 billion in damages

The trial comes as OpenAI prepares for a potential initial ​public offering that could value it at $1 trillion, Reuters has reported.

The company also faces ​growing competition from ⁠rivals, including Anthropic, while a Wall Street Journal report that OpenAI had missed some internal performance targets weighed on the tech-heavy Nasdaq Composite on Tuesday.

Musk is seeking $150 billion in damages from OpenAI and Microsoft, with any award going to OpenAI’s charitable arm.

He also wants ⁠OpenAI to ​revert to a nonprofit, with Altman and Brockman removed as officers and Altman ​removed from the board.

His claims include breach of charitable trust and unjust enrichment.

OpenAI is currently structured as a public benefit corporation, in which the nonprofit and ​other investors, including Microsoft, hold stakes.

The charges under the Digital Services Act, which requires Big ​Tech to do more to tackle illegal and harmful ​content on their platforms, came after a two-year ⁠long investigation by the European Commission.

The EU tech enforcer ​said Meta does not do enough to enforce its restrictions ​against children under 13 from using Facebook and Instagram and that measures to identify and remove them when they do access the services ​were inadequate.

It said 10-12% of children under 13 in ​Europe use Facebook and Instagram.

“Our preliminary findings show that Instagram and ‌Facebook ⁠are doing very little to prevent children below this age from accessing their services,” EU tech chief Henna Virkkunen said in a statement.

“Terms and conditions should not be mere ​written statements, but ​rather the ⁠basis for concrete action to protect users – including children,” she said.

The Commission said both platforms ​must change their risk assessment methodology and ​that they ⁠need to strengthen measures to prevent, detect and remove minors from their services.

Meta can respond to the charges and ⁠take measures ​before the Commission issues a ​decision. DSA breaches can cost companies fines as much as 6% of their ​global annual turnover.

The world’s richest person is suing OpenAI, its co-founder and Chief Executive Sam Altman and its President Greg Brockman, saying they betrayed him and the public ‌by abandoning OpenAI’s mission to be a benevolent steward of AI for humanity, and transforming the nonprofit into a profit-seeking juggernaut.

“If we make it OK to loot a charity, the entire foundation of charitable giving in America will be destroyed,“ Musk testified on the first day of the trial. “That’s my concern.”

Musk, who founded automaker Tesla and rocket company SpaceX, characterised OpenAI as his brainchild as well.

“I came up with the idea, the name, recruited the key people, taught them everything I know, provided all of the initial funding,” Musk said.

“It was ​specifically meant to be for a charity that does not benefit any individual person. I could’ve started it as a for profit and I specifically chose not to.”

Before Musk began testifying, William Savitt, a lawyer for ​OpenAI and Altman, told jurors during his opening statement that it was Musk who saw dollar signs as he helped finance OpenAI’s early growth and pushed it to become ⁠a for-profit business, one he might eventually lead as CEO.

Savitt said Musk wanted “the keys to the kingdom,” and sued only after he failed.

In 2023, he started his own AI business, xAI, now part of SpaceX.

“What he cares about is Elon ​Musk being on top,” Savitt said in his opening statement. “We are here because Mr. Musk didn’t get his way.“

OpenAI’s lawyer also framed OpenAI’s March 2019 creation of a for-profit entity as critical to letting it buy computing power and pay ​top scientists to stay competitive with Google’s DeepMind AI lab.

Musk’s lawyer, Steven Molo, told jurors in his opening statement that it was the OpenAI defendants who were greedy for money, as OpenAI began drawing investors, including Microsoft, which invested $10 billion in January 2023.

“It wasn’t a vehicle for people to get rich,” Molo said.

Musk is expected to resume his testimony on Wednesday.

Musk admonished for social media use

Musk is seeking $150 billion in damages from OpenAI and Microsoft, one of its largest investors, with proceeds going to OpenAI’s charitable arm.

He also wants OpenAI ​to revert to a nonprofit, with Altman and Brockman removed as officers and Altman removed from its board.

Musk’s claims include breach of charitable trust and unjust enrichment.

While Musk described OpenAI as a charity, the organisation called itself a nonprofit ​artificial intelligence research company in a 2015 post, “Introducing OpenAI.”

Before jurors were seated, US District Judge Yvonne Gonzalez Rogers admonished Musk after OpenAI lawyers complained about his posts on X on Monday, in which he assailed Altman as “Scam Altman” and accused him of stealing ‌a charity.

Rogers said ⁠she was loath to issue a gag order, but urged Musk to “try to control your propensity to use social media to make things work outside the courtroom … Perhaps you’ve never done that before.”

Musk agreed to minimise his social media activity, and Altman similarly agreed.

Altman and Microsoft chief Satya Nadella are also expected to testify.

The trial offers a window into some of the egos and personalities that shaped OpenAI as it evolved from a nonprofit research lab in Brockman’s apartment to a company worth more than $850 billion.

It also risks complicating OpenAI’s plans for a potential initial public offering by casting doubt on its leadership, and could intensify Americans’ fears about AI technology more broadly.

Importance of AI safety

Musk ​and Altman co-founded OpenAI in 2015 with a goal ​of developing AI to benefit humanity and fend off ⁠rivals such as Google.

Musk testified that “I’ve had extreme concerns about AI for a very long time,” and focused more intently on it after meetings with former US President Barack Obama and Google didn’t address AI’s risks.

“I was very close friends with Larry Page at Google,” Musk testified, referring to Google’s co-founder.

“We would talk for many hours about AI safety. ​At a certain point, it was clear to me Larry Page was not sufficiently caring about AI … We had to have a counterpoint against Google.”

Savitt, in his ​opening statement, said AI safety wasn’t ⁠a priority for Musk, and that Musk denigrated OpenAI employees who focused on it. “Jackasses is what he called them,” Savitt said.

Musk has said he provided about $38 million to OpenAI for its original mission, and testified he flexed his connections to provide computing capacity, personally approaching Nadella as well as Nvidia CEO Jensen Huang.

OpenAI created its for-profit entity 13 months after Musk left its board.

Russell Cohen, a lawyer for Microsoft, said in his opening statement that the company didn’t do anything wrong and has been “a ⁠responsible partner every ​step of the way.”

OpenAI also faces growing competition from rivals, including Anthropic, and is spending billions on computational resources. 

A potential IPO could value ​the company at $1 trillion, Reuters has reported.

Musk’s xAI trails far behind OpenAI in usage. He has folded that business into SpaceX, whose own potential IPO this year could be the largest ever.

Late last year, OpenAI overhauled its structure again to become a public benefit corporation, in which the nonprofit and ​other investors, including Microsoft, hold stakes.

The nonprofit holds a 26% stake, plus warrants if OpenAI hits certain valuation targets.

“The defendants in the case stole a charity, and we’re asking you ‌to hold them accountable,” Musk’s lawyer, Steven Molo, told jurors in his opening statement in an Oakland, California, federal court.

Musk is suing OpenAI, its Chief Executive Sam Altman and its President Greg Brockman, saying they betrayed him and the public by abandoning the ChatGPT maker’s mission to be a benevolent steward of AI for humanity, and transforming the nonprofit into a profit-seeking juggernaut.

A defence lawyer will make an opening statement later on Tuesday. Musk, Altman and Brockman attended the trial.

Musk, the world’s richest person, ​is seeking $150 billion in damages from OpenAI and Microsoft, one of its largest investors, with proceeds going to OpenAI’s charitable arm.

He also wants OpenAI to revert to a nonprofit, with Altman and Brockman ​removed as officers and Altman removed from its board. Musk’s claims include breach of charitable trust and unjust enrichment.

Even before jurors were seated, there was conflict as OpenAI ⁠lawyers complained about Musk’s posts on X on Monday, in which he assailed Altman as “Scam Altman” and accused him of stealing a charity.

US District Judge Yvonne Gonzalez Rogers said she was loath to issue a gag ​order, and urged Musk to “try to control your propensity to use social media to make things work outside the courtroom … Perhaps you’ve never done that before.”

Musk agreed to minimise his social media activity, as did Altman.

SEED MONEY

OpenAI ​was co-founded by Musk and Altman in 2015 with a goal of developing AI to benefit humanity and fend off rivals such as Google.

Molo said “Elon became more worried” as the technology advanced, and collaborated with Altman to “develop AI safely” after a meeting with US President Barack Obama in 2015 did not address AI’s risks.

“It wasn’t a vehicle for people to get rich,” Molo said.

Molo said Musk eventually recruited top AI scientists like Ilya Sutskever, contributed seed funding to OpenAI, and leveraged relationships ​to strike a partnership with Microsoft.

“He developed a strategy. He taught them all he knows about building a business,” Molo said. “Without Elon Musk, there would be no OpenAI.”

Musk, the Tesla and SpaceX founder, has said ​he provided about $38 million to OpenAI for its original mission, only to see OpenAI create a for-profit entity in March 2019, a little over a year after he left its board.

OpenAI countered that Musk knew about and supported the ‌transformation, and sued ⁠only after failing to become CEO, and starting his own AI company to stunt its growth.

Molo said it’s okay for a non-profit to set up a for-profit entity if they share a mission, likening it to a museum opening a gift shop.

“The museum store can’t loot the museum and sell the Picassos,” he said, later adding: “To steal a charity is absolutely wrong.”

Molo said a major turning point for Musk came when Microsoft invested $10 billion in OpenAI in January 2023, valuing the latter at $20 billion.

“This was not consistent with the nonprofit’s mission,” he said. “It violated every commitment (the defendants) made, not just to Elon, but to the world.”

Musk, Altman and Microsoft chief ​Satya Nadella are among the witnesses expected to ​testify.

Rogers has said she wants jurors to begin ⁠deliberations on the defendants’ liability by May 12. If they find the defendants liable, both sides will argue possible remedies to the judge.

EGOS AND PERSONALITIES

The trial could offer a window into some of the egos and personalities that shaped OpenAI as it evolved from a nonprofit research lab in Brockman’s apartment to a company worth ​more than $850 billion.

It also risks complicating OpenAI’s plans for a potential initial public offering by casting doubt on its leadership, and could intensify Americans’ fears about ​AI technology more broadly.

OpenAI has ⁠argued Musk was motivated by jealousy in trying to undermine its growth and prop up his own xAI, which he founded in 2023 shortly after OpenAI launched ChatGPT.

It has said Musk was involved in discussions to create OpenAI’s new structure and demanded to be CEO.

Microsoft has denied having colluded with OpenAI and says it teamed up with OpenAI only after Musk left.

OpenAI also faces growing competition from rivals including Anthropic, and is spending billions on computational resources. A ⁠potential IPO could ​value the company at $1 trillion, Reuters has reported.

Musk’s xAI trails far behind OpenAI in usage. He has folded that business into his ​rocket company SpaceX, whose own potential IPO this year could be the largest ever.

In the fall, OpenAI overhauled its structure again to become a public benefit corporation, in which the nonprofit and other investors including Microsoft hold stakes. The nonprofit holds a 26% stake, plus ​warrants if OpenAI hits certain valuation targets.

A public benefit corporation could make OpenAI more investor-friendly while retaining its charitable origins.

Addressing a ceremony in Islamabad, Special Assistant to the Prime Minister on Climate Change Romina Khurshid Alam said modern and accessible technology is the need of the hour in view of climate risks.

The special feature of Noor is that it will provide timely information to mobile phone users without internet, which will directly benefit people living in remote areas.

The project has been launched with the support of the UK’s International Development Agency.

The agreement allows the Pentagon to use Google’s AI for “any lawful government purpose”, the report added, ​putting it alongside OpenAI and Elon Musk’s xAI, which also have deals to supply ​AI models for classified use.

Classified networks are used to handle a ⁠wide range of sensitive work, including mission planning and weapons targeting.

The Pentagon signed ​agreements worth up to $200 million each with major AI labs in 2025, including Anthropic, OpenAI, ​and Google.

The Pentagon is seeking to preserve all flexibility in defence and not be limited by warnings from the technology’s creators against powering weapons with unreliable AI.

Google’s agreement requires ​it to help in adjusting the company’s AI safety settings and filters at ​the government’s request.

The contract includes language noting “the parties agree that the AI System is not ‌intended ⁠for, and should not be used for, domestic mass surveillance or autonomous weapons (including target selection) without appropriate human oversight and control”, according to the report, but also adds that the “Agreement does not confer any right to control or veto lawful ​Government operational decision-making”.

Reuters ​could not verify ⁠the report. Alphabet and the US Department of Defence, which has now been renamed the Department of War by President ​Donald Trump, did not immediately respond to requests for comment.

A spokesperson ​for Google ⁠Public Sector, the unit that handles US government business, told The Information that the new agreement is an amendment to its existing contract.

Reuters had earlier reported that the ⁠Pentagon had ​been pushing top AI companies such as OpenAI ​and Anthropic to make their tools available on classified networks without the standard restrictions they apply ​to users.

The reworked tie-up, announced jointly by the companies on Monday, retains Microsoft as OpenAI’s primary cloud partner with a licence to the startup’s intellectual property through 2032.

But ​the software giant will no longer share revenue for the OpenAI products it sells on its cloud. Revenue OpenAI must share with ​Microsoft through 2030 will now have a cap for the total number and no longer tied to the startup’s ⁠technology milestones - including if it achieves artificial general intelligence, the point at which AI matches or surpasses human ability.

The change is meant to simplify a ​complex relationship between OpenAI and its one of its biggest and earliest backers.

Microsoft’s early bet on OpenAI allowed the company to roll out AI across its ​products and powered sales growth at its Azure cloud-computing business, turning the company into one of one the biggest players in the high-stakes race for the technology.

But tensions have been rising between the companies as OpenAI strikes cloud deals with rival providers to secure more computing power and build out an enterprise business that can compete better with Anthropic ​ahead of a potential IPO.

The Financial Times reported last month Microsoft was weighing legal action against Amazon and OpenAI over a $50 billion cloud deal that may ​breach its exclusive cloud tie-up.

In an internal memo reported by CNBC, OpenAI said that the Microsoft partnership had been foundational but had limited the startup’s enterprise reach, adding ‌that demand ⁠since OpenAI launched on Amazon’s cloud had been staggering.

“The new deal with Microsoft was essential for OpenAI to be successful in the enterprise market,” said Gil Luria, analyst at D.A. Davidson & Co.

“AWS and Google Cloud enterprise customers have been limited in their ability to integrate OpenAI’s products because of the exclusive relationship and will now be more likely to consider OpenAI alongside Anthropic,” he added.

MICROSOFT WORKS TO REDUCE OPENAI RELIANCE

Microsoft shares initially fell 1.3% on the news ​but were trading little changed by ​late morning trading.

Alphabet and Amazon both ⁠did not immediately respond to Reuters’ requests for comment. Alphabet shares were up 1.5%, while Amazon was down 0.5%.

Microsoft and OpenAI had also announced restructured their tie-up in October, removing major constraints on the startup’s ability to raise capital ​and secure computing resources.

The software giant has, in recent months, been working to reduce its reliance on OpenAI ​by developing its own ⁠AI models and rolling out those developed by the likes of Anthropic in its products, including the 365 Copilot for enterprises.

It has also been said that it has been constrained by AI capacity, which has limited growth for its cloud business.

“From Microsoft’s perspective, it does not need to build out all the data centre ⁠needs for ​OpenAI, freeing up capital for Copilot and other cloud capacity,” Barclays analysts said, calling the ​move a positive for both Microsoft and OpenAI.

Ending the exclusivity pact may help Microsoft fight antitrust scrutiny in the UK, the US and Europe over whether its OpenAI tie-up gives it an unfair advantage in the ​cloud and enterprise AI markets.

The research, published in Nature and highlighted by ScienceDaily, presents a new model of human origins. It suggests that early populations of Homo sapiens were spread across the African continent and remained genetically connected over long periods.

Instead of a simple evolutionary “family tree,” scientists propose a complex web of populations that split, migrated and reconnected over time, exchanging genetic material.

The study combined genomic data from present-day African populations with fossil evidence of early humans. One key component was the analysis of 44 newly sequenced genomes from the Nama people of southern Africa, known for their high genetic diversity.

Lead author Brenna Henn noted that limited fossil records and ancient DNA samples have long made it difficult to fully understand early human history. However, the new findings significantly reshape the narrative.

Co-author Tim Weaver said the results highlight the need to rethink traditional models of human evolution, emphasising interconnected populations rather than isolated lineages.

Scientists broadly agree that modern humans originated in Africa. However, this study underscores that the process was far more complex than previously thought — involving continuous interaction among diverse groups rather than a single point of origin.

The cable, dated Friday and sent to diplomatic and consular posts around the world, instructs diplomatic staff to speak to their foreign counterparts about “concerns over adversaries’ extraction and distillation of US AI models.”

“A separate demarche ​request and message has been sent to Beijing for raising with China,” the document states.

Distillation is the process of training smaller ​AI models using output from larger, more expensive ones as part of an effort to lower the ⁠costs of training a powerful new AI tool.

This week, the White House made similar accusations, but the cable has not been previously reported.

The ​State Department did not immediately respond to a request for comment.

OpenAI has warned US lawmakers that DeepSeek was targeting the ChatGPT maker and the ​nation’s leading AI companies to replicate models and use them for its own training, Reuters reported in February.

China rejects accusations

The Chinese Embassy in Washington on Friday reiterated its stance that the accusations are baseless.

“The allegations that Chinese entities are stealing American AI intellectual property are groundless and are deliberate attacks on China’s development and ​progress in the AI industry,” it said in a statement to Reuters.

DeepSeek, whose low-cost AI model stunned the world last year, on Friday launched ​a preview of a highly anticipated new model, called the V4, adapted for Huawei chip technology, underlining China’s growing autonomy in the sector.

DeepSeek also did ‌not immediately ⁠respond to a request for comment.

In the past, it has said that its V3 model used data naturally occurring and collected through web crawling and it had not intentionally used synthetic data generated by OpenAI.

Many Western and some Asian governments have banned their institutions and officials from using DeepSeek, citing data privacy concerns.

Nevertheless, DeepSeek’s models have consistently been among the most used on international platforms that host open-source models.

The ​State Department cable said its purpose ​was to “warn of the risks ⁠of utilising AI models distilled from US proprietary AI models, and lay the groundwork for potential follow-up and outreach by the US government.”

It also mentioned Chinese AI firms Moonshot AI and MiniMax.

Neither company immediately ​responded to a request for comment.

The cable said that “AI models developed from surreptitious, unauthorised distillation campaigns ​enable foreign actors to ⁠release products that appear to perform comparably on select benchmarks at a fraction of the cost but do not replicate the full performance of the original system.”

It added that the campaigns also “deliberately strip security protocols from the resulting models and undo mechanisms that ensure those AI models are ideologically ⁠neutral and ​truth-seeking.”

The White House accusations and the cable come just weeks before US President ​Donald Trump is set to visit Chinese President Xi Jinping in Beijing.

They could well raise tensions in a long-running tech war between the rival superpowers, which had been ​lowered by a detente brokered last October.

Like several global automakers, Honda was forced to quickly adjust its strategy in mid-2025 after a sudden policy change in the United States removed a long-standing EV tax credit.

The sudden policy change dealt a heavy blow to the industry.

Ford and General Motors both posted losses running into the billions, while Honda’s losses were even steeper, reportedly topping $15.7 billion.

Last year, Mibe admitted that changes in US policy may have slowed the transition away from internal combustion engines, although he stressed Honda’s responsibility to help reduce global emissions by moving toward cleaner technologies.

Following the company’s first annual loss, revealed in early March shortly after his China visit, Mibe expressed concern over the efficiency of the EV facility he toured.

“We have no chance against this,” he was quoted as saying in a Nikkei Asia report, describing the plant’s highly automated operations where production, logistics, and supply chains were fully digitised with minimal human presence on the factory floor.

Online reactions to the wider EV debate were divided.

While many users said there is a clear demand from consumers, but not enough supply to meet it, others pointed to people switching vehicles and changing brand loyalties.

Many users stepped back to look at the broader picture, saying they were worried the US could fall behind in the global move toward cleaner energy if policies keep shifting.

They added that this lack of consistency could end up hurting the economy over the long run.

Opening statements in Musk’s civil lawsuit against OpenAI and Altman will take place in the Oakland, California, federal court, following the selection on Monday of nine jurors.

Musk claims that Altman and Greg Brockman, respectively OpenAI’s chief executive and president, betrayed him and the public by abandoning the company’s mission to be a benevolent ​steward of AI for the benefit of humanity, and turning it into a “wealth machine” for themselves and investors.

The world’s richest person is ​seeking $150 billion in damages from OpenAI and Microsoft, one of its largest investors, with proceeds going to OpenAI’s charitable ⁠arm.

He also wants OpenAI to revert to a nonprofit, with Altman and Brockman removed as officers and Altman removed from its board.

Musk, the Tesla and ​SpaceX founder, has said he provided about $38 million of seed money to OpenAI for its original mission, only to see OpenAI create a for-profit entity ​in March 2019, a little over a year after he left its board.

OpenAI countered that Musk knew about and supported the transformation, and sued only after failing to become CEO, and starting his own AI company to stunt its growth.

Musk is no longer seeking damages for himself as he pursues breach of charitable trust and unjust enrichment ​claims.

US District Judge Yvonne Gonzalez Rogers has said she wants jurors to begin deliberations on the defendants’ liability by May 12.

The jurors include nurses, ​city workers and retirees. If they find the defendants liable, both sides will argue possible remedies to the judge.

Musk, Altman and Microsoft chief Satya Nadella are among the ‌witnesses expected ⁠to testify, with Musk taking the stand as soon as this week.

Egos and personalities

Musk and Altman co-founded OpenAI in 2015 with a goal of developing AI to benefit humanity and fend off rivals such as Google.

The trial could offer a window into some of the egos and personalities that shaped OpenAI as it evolved from a nonprofit research lab in Brockman’s apartment to a company worth more than $850 billion.

It risks complicating OpenAI’s plans for a ​potential initial public offering by casting doubt ​on its leadership, and could ⁠also intensify Americans’ fears about AI technology more broadly.

OpenAI has argued that Musk was motivated by jealousy in trying to undermine its growth and prop up his own xAI, which he founded in 2023 shortly after OpenAI ​launched ChatGPT.

It has been said that Musk was involved in discussions to create OpenAI’s new structure and demanded to be CEO.

Microsoft ​has denied having colluded ⁠with OpenAI and says it teamed up with OpenAI only after Musk left.

OpenAI faces growing competition from rivals, including Anthropic, and is spending billions on computational resources. A potential IPO could value the company at $1 trillion, according to Reuters.

Musk’s xAI trails far behind OpenAI in usage. He has folded that business into his ⁠rocket company ​SpaceX, whose own potential IPO this year could be the largest ever.

Last fall, OpenAI overhauled ​its structure again to become a public benefit corporation, in which the nonprofit and other investors, including Microsoft, hold stakes. The nonprofit holds a 26% stake, plus warrants if OpenAI hits ​certain valuation targets.

A public benefit corporation could make OpenAI more investor-friendly while retaining its charitable origins.

The initiative, intended to modernise classrooms and reduce inequality, was instead described by former governor Paul LePage as a “massive failure,” after evidence showed no measurable rise in academic performance.

Reports suggest the impact also varied by region.

In wealthier school districts, students reportedly used laptops for creative and collaborative work, while in poorer and rural areas, they were largely limited to basic tools such as Word and PowerPoint, widening digital disparities rather than reducing them.

The programme is part of a broader trend across the US, where spending on education technology reached an estimated $30 billion in 2024, far exceeding textbook budgets.

Despite this investment, analysts cited in the report say student outcomes have not improved in line with spending.

Studies referenced in the reporting also point to rising concerns over classroom screen use, with teachers estimating that students spend several hours a day on digital devices during lessons.

Research has found that a significant share of laptop use in educational settings is unrelated to coursework, with distractions such as social media dominating screen time.

Experts say the findings raise wider questions about the effectiveness of large-scale digital learning strategies, particularly as concerns grow over declining academic performance among younger generations in some Western countries.

Education researchers argue the issue is not technology itself, but how it is used, noting that devices designed for learning often compete with highly engaging entertainment platforms.

The AI race has intensified the rivalry between China and the United States, with the White House on Thursday accusing Chinese entities of a massive effort to steal artificial intelligence technology. Beijing called the claim “baseless”.

Hangzhou-based DeepSeek burst onto the scene in January last year with a generative AI chatbot, powered by its R1 reasoning model, that upended assumptions of US dominance in the strategic sector.

DeepSeek-V4 “features an ultra-long context”, the company said in a statement on social media platform WeChat, hailing it as “world-leading… with drastically reduced compute (and) memory costs” in a separate announcement on X.

V4 supports a context length of one million “tokens” – small components of text, including words or punctuation — putting it on par with Google’s Gemini.

Context length determines how much input a model is able to absorb to help it complete tasks.

The new V4 is released as two versions, DeepSeek-V4-Pro and DeepSeek-V4-Flash, with the latter being “a more efficient and economical choice” because it has smaller parameters.

In terms of “world knowledge”, a benchmark for reasoning, V4-Pro trails only the latest Gemini model, DeepSeek said.

A “preview version” of the open source model is now available, the company said, without indicating when a final version would be released.

‘Inflection point’

Experts say V4’s arrival marks an “inflection point” in terms of hardware and cost.

“This addresses the long-standing issues of slower performance and higher costs associated with long context lengths, marking a genuine inflection point for the industry,” Zhang Yi, the founder of tech research firm iiMedia, told AFP.

“For end users, this will bring widespread, accessible benefits. For instance, if ultra-long context support becomes a standard feature, long-text processing is expected to move beyond high-end research labs and enter mainstream commercial applications,” he said.

V4-Pro has 1.6 trillion parameters while the V4-Flash has 284 billion parameters, which refine models’ decision-making ability.

The model has also been “optimised” for popular AI Agent products such as Claude Code, OpenClaw, OpenCode and CodeBuddy, the DeepSeek statement said.

It can also run on chips manufactured by Chinese tech giant Huawei, the company added.

Huawei — sanctioned by the US since 2019 over national security — said in a statement Friday that the full range of its Ascend SuperPoD products are supporting DeepSeek’s V4 series.

DeepSeek’s latest release is a “milestone” for Chinese firms, said veteran AI industry analyst Max Liu.

“It’s a good thing for the entire domestic AI industry. It can provide better models for domestic users and we can now expect a lot more things – more products (and a) more competitive market,” he told AFP.

“This is no less shocking than when DeepSeek first came out” if its new model indeed matches the performance of leading models from Western labs, he added.

‘Sputnik moment’

Last year’s so-called “DeepSeek shock” sparked a sell-off of AI-related shares and a reckoning on business strategy in what was also described as a “Sputnik moment” for the industry.

The chatbot performed at a similar level to ChatGPT and other top American offerings, but the company said it had taken significantly less computing power to develop.

However, its sudden popularity raised questions over data privacy and censorship, with the chatbot often refusing to answer questions on sensitive topics such as the 1989 Tiananmen crackdown.

DeepSeek’s AI tools have been widely adopted by Chinese municipalities and healthcare institutions as well as the financial sector and other businesses.

This has been partly driven by DeepSeek’s decision to make its systems open source, with their inner workings public — in contrast to the proprietary models sold by OpenAI and other Western rivals.

But the White House has accused Chinese firms of vying to “steal” American technology, ahead of an expected summit between Donald Trump and Xi Jinping in Beijing next month.

“The US has evidence that foreign entities, primarily in China, are running industrial-scale distillation campaigns to steal American AI,” Trump’s science and technology chief adviser Michael Kratsios said in a post on X.

Distillation is a common practice within AI development, often used by companies to create cheaper, smaller versions of their own models.

“The US claims are entirely baseless,” Chinese foreign ministry spokesman Guo Jiakun told a news conference in Beijing. “They are a slanderous smear against the achievements of China’s artificial intelligence industry.”

Rows of influencers posed in front of gleaming models at the capital’s cavernous international exhibition centre, darting to suitcases stuffed with outfit changes, while animated CEOs worked the crowds in front of massive LED screens.

Legacy overseas brands such as Volkswagen, Toyota and BMW once dominated in China, but have lost market share in past years to domestic firms that beat them to the electric vehicle revolution and undercut them on price.

Chinese manufacturers including BYD, Xiaomi and XPeng are now also at the forefront of integrating AI software and autonomous driving technology into their EVs.

The Auto China exhibition, hosted at two side-by-side venues, spans 380,000 square metres (four million square feet), according to organisers — more than 50 football pitches.

More than 1,400 vehicles from hundreds of Chinese and foreign companies are on display from Friday, when the show opened to industry professionals and the media, and later to the public from April 28 until May 3.

While traditional leading brands like Germany’s BMW and Mercedes held sweeping areas of the vast halls, most of the event’s mega stages were dominated by Chinese brands, including BYD and battery giant CATL.

Robots, flying cars

At the sprawling expo, crowds cheered as XPeng chief executive He Xiaopeng unveiled the company’s new GX, a six-seat electric SUV.

The imposing 5.2-metre vehicle incorporates AI technology and is aimed at breaking into the luxury market, He said.

It would soon be followed by humanoid robots this year, He promised, and eventually by flying cars, which XPeng hopes to mass-produce.

Foreign automakers are increasingly collaborating with Chinese firms to stay competitive.

BMW has partnered with CATL, while Audi is using Huawei’s driving assistance systems and Volkswagen is co-developing EVs with Guangzhou-based XPeng.

XPeng President Brian Gu said companies were “leveraging their respective strength to collaborate with China” a trend he said would continue.

Gu has his eyes set on export markets including the Gulf and Europe, where he anticipates growth to accelerate.

“In the last year, we started local production, and this year we are going to launch even more new products for the European market” which he said was responsible for half of XPeng’s global sales in 2025.

Asked by AFP how Trump’s tariffs were affecting XPeng, he said only that the US market remained an important one.

Fierce competition

This year, companies are also jostling to sell space, analysts say, with roomy SUVs’ new growth area targeting customers prioritising seating and comfort.

China “has become a customer retention and replacement/upgrade-driven market, and these big SUVs address that need,” independent analyst Lei Xing wrote in a blog this week.

Firms have flooded the domestic market in recent years with trade-in schemes, offering huge discounts to customers to give up their old car for a new one.

The fierce price war led Chinese officials last year to call for tighter price monitoring and improving long-term regulation of competition.

But newcomers appear unfazed, Lei wrote, naming at least eight EV brands from Chinese automakers that have cropped up over the last two years.

Electric vehicles, an area China dominates, are also gaining traction as rising global oil prices linked to the Middle East war push drivers away from fossil fuel-powered models.

Chinese tech on display on Friday went beyond the road.

Dozens of people queued to clamber into an enormous air taxi, a 10-seater from Chinese aviation startup AutoFlight, part of China’s wider push to dominate the low-attitude economy.

For Chinese auto enthusiast Dai, domestic EVs were the expo’s clear main characters.

“In comparison, foreign brands seem to have a weaker presence and less visibility,” the 30-year-old influencer, who gave only his surname, said.

China’s ‌most recent five-year plan, released earlier this year, presented a blueprint for “AI Plus,” a national project to embed AI systems into manufacturing, healthcare and almost every other corner of the economy.

Part of that aim is to break China’s dependence on high-end semiconductors - a trade chokepoint dominated by the US.

“There’s no longer a distinction between a technology company and a car company,” Nissan Motor China chief Stephen Ma ​told reporters on the sidelines of the Beijing Auto Show, which kicked off on Friday. “The AI-developed vehicle is much faster, and it’s quicker in China.”

In recent ​days, Chinese automakers and their suppliers have flooded the zone with investment commitments and new AI systems. Some of the immediate applications ⁠seemed incremental. Analysts say the longer-term stakes are huge.

China’s automakers are now so advanced they are upending the global car industry, said Francois Roudier, secretary general of the ​International Organisation of Motor Vehicle Manufacturers, a federation of trade groups that represents the world’s auto industry.

“There is no transition,” Roudier told Reuters in Beijing. “It’s a revolution.”

The car is the agent

Xpeng has said its updated AI model allows drivers to give the car commands - like, “park near the entrance to the shopping centre” - rather than designating a spot on a map. Xpeng vehicles can use cameras to navigate even without mapping or coordinates.

Xiaomi, an appliance and phone maker that stormed into the EV business three years ago, released an updated AI model just after midnight on Thursday.

Xiaomi has ​said its AI-empowered HyperOS operating system in its cars would allow drivers to task the system with complicated to-do lists, making restaurant reservations, placing coffee orders and compiling ​notes from the road. The system could also detect when drivers seem stressed or agitated and adjust the lighting and music for their arrival at home.

“So much focus on AI in other ‌parts of ⁠the world has been on how we can use it to improve business? That’s not what Chinese automakers are talking about,” said Dan Hearsch, global co-leader for automotive at advisory firm AlixPartners. “The AI they’re building in is going to make the car easier to drive, easier to interact with, easier to do all of the things that otherwise take effort.”

Huawei, which has pivoted from its traditional focus on telecommunications to develop businesses in chips, AI and connected cars, said it would invest more than $10 billion over the next five ​years to boost computing power for smart ​driving.

While automotive sales make up a relatively ⁠small part of Huawei’s portfolio, it remains the company’s fastest-growing segment.

Just before the auto show started, Horizon Robotics, a Chinese chipmaker that competes with Qualcomm, launched its Starry 6 processor that integrates cockpit and driving functions with the ability to handle up to ​12 screen displays in a vehicle.

A number of Chinese EV companies have been chasing Tesla by designing their own chips to ​reduce their reliance on ⁠Nvidia. That includes Xpeng, Li Auto, BYD, Geely and Leapmotor.

NIO, which spun off its chip unit, sees developing its own semiconductors as a way to reduce costs and boost earnings by swapping out Nvidia, CEO William Li said.

“We are open to the whole industry, and we welcome them to use (our chips),” Li told Reuters.

Some automakers used the Beijing auto show to ⁠demonstrate that they ​heard Beijing’s message on strategic innovation loud and clear. Dongfeng Motor - one of the Big Four state-owned ​carmakers - said it would be building cars using “embodied AI technology” in line with China’s long-term plans for the sector.

Dongfeng has been working with Huawei on smart driving systems to compete with privately owned rivals.

“When the nation ​calls, Dongfeng answers,” Chairman Yang Qing said.