OpenAI CEO Sam Altman announced plans to ease some of ChatGPT’s safety restrictions, aiming for friendlier, more “human-like” interactions. Starting in December, “verified adults” will be allowed to engage in erotic conversations, reflecting OpenAI’s “treat adult users like adults” principle. Altman acknowledged that previous restrictions aimed at mental health concerns had made the chatbot less enjoyable for many users.
This move comes amid reports of troubling relationships some users have developed with ChatGPT, including cases that led to mental health crises. OpenAI asserts it has mitigated these issues but has offered little evidence to support this claim.
OpenAI has introduced several safety features, including an age prediction system and parental controls for minors, alongside forming a mental health expert council. However, concerns persist about how the new erotic features may affect vulnerable users.
While ChatGPT now boasts 800 million weekly active users, it faces competition from other companies like Google and Meta, which heightens the pressure to attract more users. Altman stated that OpenAI’s adjustments aim to broaden appeal while balancing user safety, but the challenge of protecting vulnerable users amidst aggressive growth remains a significant concern.
Source link
