OpenAI to relax ChatGPT restrictions; plans to roll out new version in weeks

OpenAI chief executive Sam Altman said Tuesday that the company will relax some of the restrictions on its ChatGPT product, allowing users greater control over tone, personality, and access to adult material.

In a statement posted on X, Altman said ChatGPT had been made “pretty restrictive” to avoid risks associated with mental health and emotional dependency. He said OpenAI has since developed new tools to mitigate those risks and is ready to “safely relax the restrictions in most cases.”



Further, the executive confirmed that a new version of ChatGPT is expected to be released in the coming weeks.

It will let users choose how expressive or human-like they want the system to be, including options for tone, use of emoji, and conversational style, adding that in December, the company will also expand its age-verification features.

Parental controls and age restrictions

In 2024, the company released GPT-4o, a multimodal model capable of handling text, images, and real-time voice, but the company later limited those qualities after concerns were raised about users forming emotional attachments to the AI.

OpenAI recently introduced teen-specific safeguards focussed on safety, freedom, and privacy, while imposing age restrictions on sensitive queries such as requests for suicide notes or mental health advice.

CEO Sam Altman added that in some cases, the company might ask for a user ID. And, if an under-18 user has suicidal ideation, the company will attempt to contact the user’s parents and, if unable, will contact the authorities in case of imminent danger.

For teens, the company said safeguards to “potential serious misuse” and queries that are disruptive in nature, causing harm to someone’s life or society at large, would be escalated for human review.