Summary
- OpenAI has indefinitely canceled the release of Adult Mode, a feature that would have allowed the creation of erotic content and conversations.
- The age prediction algorithm presented rate 12% error, incorrectly identifying minors as adults, which caused strong reactions.
- Scientific advisors warned of the risk of serious emotional dependence among users.
- The company is now focusing on the core functions of the model, avoiding legal problems that competitors like xAI already face.
- In Greece and European Union, the strict application of the AI Act made this operation extremely difficult from a legal point of view.
OpenAI has indefinitely suspended the release of the so-called Adult Mode for ChatGPT, abandoning the attempt to integrate sex chat features and create NSFW (Not Safe For Work) content.
According reference According to the Financial Times, the decision was made after a lot of internal backlash, investor pressure and concerns from the scientific community. The cancellation of this feature demonstrates the company's change of strategy, which chooses to focus on improving the intelligence of its core models and avoid legal risks.
Why is OpenAI abandoning Adult Mode in ChatGPT?
OpenAI has indefinitely suspended the Adult Mode due to the weakness of its age identification algorithm, which recorded an error rate of 12%. The decision was made after safety warnings from mental health consultants and investor pressure, forcing the company to focus exclusively on improving its core productive artificial intelligence.
The technical obstacle of age identification
The main technological problem that led to the freezing of Adult Mode concerns the development of a reliable system Age Prediction (Age Prediction). ChatGPT was based on an automated process that attempted to guess the user's age by analyzing speech patterns, the general vocabulary of prompts, and the hours of use of the platform.
Test data revealed that the algorithm incorrectly classified minors as adults 12% of the time. With ChatGPT counting over 900 million users worldwide, such an error rate translates into millions of underage users who would have potentially had access to inappropriate and sexualized content. OpenAI’s security engineers were unable to reduce the margin of error to levels acceptable for commercial release, making the venture unviable.
Psychologists' warnings and legal risks
OpenAI’s internal wellness and ethics board had expressed its complete opposition to the development of this product. The main concern of cognitive scientists and psychologists was the risk of creating unhealthy emotional dependence. Given the already existing tendency of users to anthropomorphize artificial intelligence, experts argued that NSFW conversations could lead vulnerable people into dangerous psychological states.
The Wall Street Journal revealed that product policy executives openly voiced their objections, highlighting the risk of promoting a “digital escort” that would increase psychological isolation. In addition, the company’s legal department warned about the potential for the system to be used to produce non-consensual content, despite the decision to limit the feature to text-only.
The competition and the mistakes of xAI
The development comes amid a broader industry realignment around security issues. OpenAI has been following the legal blows to Elon Musk’s xAI platform Grok, which has faced lawsuits over inadequate security filters that allowed the spread of illegal material and images of minors.
OpenAI's management has decided not to follow the same dangerous path. By canceling Adult Mode (as well as limiting other experimental video production platforms), Sam Altman seems to prefer to shield the company from federal investigations and legal sanctions, focusing resources on the enterprise part, which is also the most profitable.
What does the cancellation mean for Greece and the European Union?
The European and consequently the Greek market is now strictly regulated by the AI Act, which classifies AI systems based on risk. The availability of an Adult Mode in Europe would require strict identification mechanisms (Kyc/Identity Verification) through government documents, which is in direct conflict with European data protection rules (GDPR).
Techgear's view
OpenAI's retreat on the issue of Adult Mode proves that technological limits are no longer determined solely by computing power, but by social and legal responsibility.
Text analysis for age prediction cannot replace hard-verification systems. The decision to cut these experimental "side quests" shows maturity, protecting users from the misuse of AI for malicious purposes, while maintaining ChatGPT as a fundamentally safe and productive tool.
