Sam Altman, CEO of OpenAI, is pushing to introduce an "adult mode" in ChatGPT. Altman previously stated on social media that the company would follow the principle of "treating adult users as adults," allowing verified adult users to access more diverse content, including erotica, while ensuring safety.

Although Altman had promised to relax restrictions on "mature applications" as early as five months ago, the implementation of this plan has been slow. According to the Wall Street Journal, there are intense disagreements within OpenAI about this feature:

  • Leadership Disagreements: Many employees and executives were caught off guard by Altman's sudden commitment, fearing it could lead to serious ethical and brand crises.

  • Safety Risks: OpenAI's current age prediction system has an error rate of about 12%, meaning millions of minors could bypass restrictions to access inappropriate content.

  • Mental Health Concerns: Advisors have warned that overly intimate AI interactions could lead users to develop severe emotional dependencies.

To mitigate risks and compete with rivals, OpenAI has adopted a relatively cautious strategy. Unlike xAI, owned by Musk, which is deeply involved in litigation over generating illegal images, OpenAI's adult mode will be limited to text generation. A company spokesperson emphasized that the content generated by this feature is closer to "romantic novels" rather than purely explicit material, and will encourage users to build connections in the real world.

Critics point out that OpenAI's move is largely aimed at alleviating current financial pressure by expanding content boundaries to gain more subscription revenue. However, with the previous head of security leaving the company due to opposition to the plan, and increasing regulatory scrutiny on AI-generated content, it remains uncertain whether this "adult mode" will be launched as planned within a month.