OpenAI reportedly plans social network with biometric bot detection
OpenAI is reportedly planning its own social network that will keep bots out using biometric methods. However, this raises data protection concerns.
(Image: kovop/Shutterstock.com)
They cause unrest, are intended to destabilize society, and sow panic: bots have become a major problem on social networks. For many people, they are difficult or impossible to recognize as such. The AI developer OpenAI is now reportedly planning its own social network, according to a report by the US magazine Forbes, which is intended to prevent bots from entering in the first place. Biometric methods are to be used for this verification.
For this “Proof of Personhood”, OpenAI wants to use either Apple's iPhone Face ID sensor or the World Network Iris Scanner, the report states. However, the device from Tools for Humanity is controversial; the company is also known as the issuer of the cryptocurrency WLD token. Spain and Portugal have temporarily banned iris scanning due to data protection concerns.
The Bavarian State Office for Data Protection Supervision also identified “fundamental data protection risks” in December 2024. Unlike a password, eyes cannot be changed; data protection advocates explain their reservations. World Network, conversely, complained that the EU has no clear anonymization standards. However, it is unlikely that OpenAI will forgo iris scanning: OpenAI CEO Sam Altman is both CEO of OpenAI and chairman of Tools for Humanity.
Videos by heise
Why OpenAI wants its own network
In the planned bot-free social network, OpenAI envisions only humans mingling. Whether this will solve all problems, however, is doubtful. It first became known in April 2025 that OpenAI is working on a social network prototype with a small team.
According to media reports, X alone deleted 1.7 million bot accounts in 2025. OpenAI's interest in its own social network is not only due to the problem with bots: the ChatGPT developer is facing competitors who own their own social networks and use them to make their chatbots more well-known. Above all, however, the networks are an access to data that can be used for real-time training of AI models. X has integrated Grok, for example. Meta also uses its own AI models in its networks Facebook and Instagram, as well as in the messenger WhatsApp.
(mki)
