OpenAI is quietly preparing to build a new social network product. Its core selling point is envisaged as "only for real users", trying to fundamentally solve the problem of robot accounts that have long plagued social platforms, especially X (formerly Twitter). According to people familiar with the matter, the project is still in the very early stages of development and is being promoted by a small team of less than 10 people, but internal ideas have gradually focused on using biometric technology to verify user identities.
According to reports, the team has discussed requiring users to provide some kind of "proof of personhood" (proof of personhood), one way is through Apple's Face ID, and the other is using World's "eye-scanning spherical device." Tools for Humanity, the company behind the cantaloupe-sized device that generates a unique and verifiable identity for an individual through iris scanning, was founded and is chaired by OpenAI CEO Sam Altman. If biometric verification is truly adopted, it will mean that there must be a real person behind each account on OpenAI's new platform. This is different from Facebook, LinkedIn and other practices that rely more on mobile phone numbers, email addresses or behavioral patterns to identify users.
However, biometric solutions also raise widespread privacy concerns. Privacy advocates warn that once data such as iris scans are leaked or misused, the consequences may be far more serious than ordinary account information, because biometrics themselves are immutable, and once mastered by criminals, the potential risks are difficult to reverse. It’s unclear how OpenAI will strike a balance between fighting bots and protecting privacy, and there’s no indication of which specific technology path the company has finally decided on.
In terms of product positioning, it is currently unclear how this social network will connect with OpenAI’s existing application matrix. People familiar with the matter said that it is "likely" that the platform will allow users to directly use AI to generate content, such as videos or images, to form three-dimensional linkages with applications such as ChatGPT and Sora. Similar features have already appeared on Meta's Instagram, which has about 3 billion monthly active users, and users can generate AI images directly within the app. There is no clear timetable for the social project to go online, and sources also cautioned that the product may still undergo major adjustments before it is officially released. OpenAI declined to comment on related issues; The Verge previously reported in April last year that OpenAI was developing an "X-like" social product.
Bot accounts have been a plague on social media for years, often simulating human interactions and used to drive up cryptocurrency prices or distort public opinion by amplifying hate speech and disinformation. The problem was particularly acute on X, and the situation is thought to have worsened after Elon Musk acquired Twitter, renamed it X and laid off about 80% of its staff, including a large "trust and safety" team responsible for content moderation and anti-bot work. Before the acquisition, Musk publicly "declared a war on spam bots" and led a cleanup operation in 2025, claiming to have deleted about 1.7 million bot accounts to reduce replies to spam messages. However, bots are still banned repeatedly.
Altman, who has been active in X for a long time, has also publicly expressed his dissatisfaction with the proliferation of robots on the platform many times. In September 2025, he posted on X that the atmosphere on Twitter and Reddit around AI-related topics "has become very fake and completely different from a year or two ago." In the previous days, he also cited the "dead Internet theory" (that is, the view that the Internet has been invaded by a large number of non-human activities since 2016), lamenting that there seem to be "a lot of X accounts driven by large language models."
From the perspective of product experience, OpenAI has a good record in popular applications for consumers. ChatGPT, which brought generative AI into the mainstream, reached 100 million users within two months of its launch and currently has more than 800 million users. Its AI video application Sora exceeded 1 million downloads within 5 days of its launch, and the growth rate is even considered faster than ChatGPT’s early performance. This also makes the outside world generally believe that once OpenAI officially launches a social product, its customer acquisition and communication capabilities cannot be underestimated.
Even so, OpenAI will still face extremely fierce competition if it enters the social field. Currently, the number of daily active users of Meta's Threads on the mobile terminal is equivalent to that of X, and the total number of users of the emerging platform Bluesky has exceeded 40 million. At the same time, giants such as Instagram and TikTok are competing for the position of "AI content destination" and have deeply integrated generative content into product experiences. Instagram head Adam Mosseri lamented on Threads in December that the platform’s feeds were being “filled with all kinds of synthetic content.”
Against this background, OpenAI is trying to use the two cards of "real person authentication" and "AI empowered creation" to enter the already highly saturated and problem-ridden social track. If it can significantly reduce the army of robots through biometric identification, it is expected to reshape the definition of "real users" on social platforms, but the controversy surrounding privacy and data security is bound to escalate. At present, there are still many unknowns about the specific form of this product, its launch time and whether it can truly "cure" robots' chronic diseases.