[ad_1]
OpenAI co-founder and CEO Sam Altman said that the company is planning to “safely relax” restrictions on what kinds of conversations ChatGPT can engage in, and by the end of the year will even allow adult users to have sexually explicit conversations with the AI system.
In a post on X on Tuesday, Altman wrote that “we made ChatGPT pretty restrictive to make sure we were being careful with mental health issues. We realize this made it less useful/enjoyable to many users who had no mental health problems, but given the seriousness of the issue we wanted to get this right.”
These restrictions were instituted after parents of children who committed suicide began to accuse ChatGPT of contributing to their children’s mental health crises or even helping to plan suicides. The parents of Adam Raine, a 16-year-old who committed suicide, have even sued OpenAI in an effort to compel the company to change its safety policies.
In a September blog post titled “Teen safety, freedom, and privacy,” Altman wrote that OpenAI would restrict teenage ChatGPT users from engaging in any discussions about suicide or self-harm. An earlier post, released in August, stated that OpenAI would strengthen its safeguards and content-blocking classifiers to prevent conversations that shouldn’t be allowed (such as helping someone to self-harm). If a user expresses suicidal intent, OpenAI said, ChatGPT should direct people to the suicide hotline, which is 988.
In his post on teen safety and freedom, Altman wrote that OpenAI has a policy to “treat our adult users like adults.” For example, he wrote, “the default behavior of our model will not lead to much flirtatious talk, but if an adult user asks for it, they should get it.”
On Tuesday, Altman wrote that OpenAI has developed new tools that enable the company to “mitigate the serious mental health issues,” and will begin relaxing ChatGPT’s content restrictions. He wrote that in the next few weeks, OpenAI will release a new version of ChatGPT “that allows people to have a personality that behaves more like what people liked about 4o,” referring to its AI model, GPT-4o.
After releasing GPT-5 in August, OpenAI removed GPT-4o from its lineup of available models on ChatGPT. This led to an outcry from ChatGPT users who developed a fondness for 4o’s personality. Eventually, OpenAI added 4o back to the lineup for paid subscribers. “If you want your ChatGPT to respond in a very human-like way, or use a ton of emoji, or act like a friend,” Altman wrote, “ChatGPT should do it.”
Altman added that in December, OpenAI will begin rolling out advanced “age-gating” systems, which will predict the age of a user based on how they use ChatGPT. Users who are found to be adults will be able to go further with ChatGPT than was previously allowed, including “erotica for verified adults.”
[ad_2]
Ben Sherry
Source link