OpenAI is quietly building a social network and is considering using biometric verification like World’s eye scan or Apple’s Face ID to ensure its users are people, not bots.
OpenAI’s development of the social network itself appears to be crystallizing around a singular mission: eliminating the bot problem that has turned the platform formerly known as Twitter into an endless, toxic pit.
Sources familiar with the project told Forbes that the social network, which is still in its early stages of development, has been conceived as a platform exclusively for real people, an attractive potential for the AI giant, which is looking to capitalize on its viral applications ChatGPT and Sora. However, if launched, it will enter a market of already consolidated and powerful platforms, such as X, Instagram and TikTok.
The app is being developed by a very small team (less than 10 people) and could include a biometric identity recognition element. Sources familiar with its development told Forbes that the team has considered requiring users to provide “proof of identity” using Apple’s Face ID or the World Orb, a melon-sized eye scanner that uses a person’s iris to generate a unique, verifiable ID. World is operated by Tools for Humanity, a company that Sam Altman, CEO of OpenAI, founded and currently chairs.
A true biometric verification would ensure that all accounts on the OpenAI social network have a real person behind them. While social networks like Facebook and LinkedIn attempt to verify identity, they typically do so through phone and email verification, or through behavioral or network signals. Neither has developed a biometric solution that, presumably, definitively proves that a user is human. Privacy advocates have warned of the risks of identity verification like World’s, as iris scans are unalterable and could cause serious problems in the wrong hands.
It has not been determined how the social network will complement OpenAI’s existing product portfolio, although some sources say that AI will likely be used to create content, such as videos or images. Meta’s Instagram, which had 3 billion monthly active users in September, now allows you to create AI-generated images directly in the application. There is currently no release date for the OpenAI social network and it could change dramatically before it is ready for public launch, sources warned.
OpenAI declined to comment. The Verge reported in April that OpenAI was working on a social network.
For years, social media has been plagued by bot accounts that often mimic human interaction to, for example, inflate cryptocurrency prices or distort public perception by amplifying hate speech. They have been a particular problem on Twitter, which worsened exponentially when Elon Musk acquired it, renamed it
Musk declared war on bots before acquiring Twitter, and in 2025, the company removed about 1.7 million bot accounts in a purge aimed at reducing response spam. However, they remain a problem.
Altman, a regular X user since 2008, has been outspoken about his frustration with bots. In September, he posted on A few days earlier, he raised a similar issue, citing the dead internet theory, which posits that since 2016, the internet has been overrun by non-human activity. «I never took the theory of the dead internet very seriously, but it seems that there are now many Twitter accounts managed by LLM [IA]”, wrote.
OpenAI has a solid track record in developing applications with great virality among consumers. ChatGPT, which popularized AI, reached 100 million users within two months of its launch and now has more than 800 million. Its AI video app, Sora, reached 1 million downloads in less than five days, a faster growth rate than ChatGPT.
Despite its track record, OpenAI will likely face an uphill battle if it decides to launch a social network. It will have to compete with Meta’s Threads app, which now has as many daily mobile users as X, and startups like Bluesky, which has more than 40 million users, not to mention giants like Instagram and TikTok, which are leading the race to become AI content destinations. “Feeds are starting to fill up with synthetic content,” Adam Mosseri, head of Instagram, posted in December.
Rich Nieva and Phoebe Liu contributed to this report.
This article was originally published in Forbes US
