This development comes as users continue to show interest in ChatGPT’s new image-generation capabilities.

The potential watermarking feature was uncovered by an X user, Tibor Blaho, who identified a line of code labeled “image_gen_watermark_for_free.” This suggests that images produced by users of the free version may soon carry watermarks, which could serve as an incentive for users to upgrade to a paid subscription.

While this is not the first instance of OpenAI considering watermarks previously, the company had developed a tool for watermarking AI-generated text there has been no official confirmation on the release of such a feature for images.

Critics have pointed out that requiring images to be watermarked could detract from user experience and interest in the product. However, if managed well, watermarking could enhance revenue prospects for OpenAI.

The design of these watermarks remains uncertain. Unlike traditional watermarks, Google’s AI image watermarking system is designed to be discreet, altering a few pixels to create a detectable pattern that doesn’t interfere with the visual integrity of the image.

Read more

Ghibli effect: ChatGPT usage hits record after rollout of viral feature

How to create ghibli-style images using Grok 3 for free not Chatgpt

Meta to rollout AI feature on WhatsApp

Introducing visible watermarks for free users while providing unmarked images to paid subscribers could create ethical questions.

If both images appear identical, the only advantage for paying users would be the ability to present AI-generated images as their own, raising concerns over authenticity.

As of now, no definitive details about the watermarking feature have been disclosed, and OpenAI may ultimately decide against its implementation.

The conversation around this feature may evolve, especially if users express their opinions to CEO Sam Altman on social media.