🔗 Share this article

The company will now limit how its AI chatbot interacts with individuals it believes are under 18, except when they pass the company’s age verification system or submit ID.

This move comes after a lawsuit from the relatives of a 16-year-old who took his own life in spring after months of exchanges with the chatbot.

Prioritizing Protection Over Freedom

CEO the OpenAI leader said in a blog post that the company is putting “safety ahead of privacy for young people,” adding that “underage users need strong protection.”

Altman explained that the system will interact in a distinct way to a 15-year-old compared to an grown-up.

Upcoming Age-Prediction Features

The AI developer aims to build an age-prediction system that estimates user age based on usage patterns. In cases where uncertainty exists, the technology will switch to the under-18 interface.

Some individuals in particular countries may also be asked to provide identification for confirmation.

“We understand this is a trade-off for adults but believe it is a necessary tradeoff.”

Enhanced Content Restrictions

For users detected to be under 18, the AI will prevent explicit material and will be trained to not engage in romantic exchanges.

Additionally, it will avoid discussions about self-harm or self-harm, including in fictional contexts.

In cases where an young user expresses thoughts of self-harm, OpenAI will attempt to notify the user’s guardians or, if not possible, alert authorities in instances of immediate danger.

Context of the Court Action

The company admitted in August that its safeguards could fall short and pledged to install more robust safety measures around harmful topics.

This action came after the parents of teenager a California youth sued the company after his passing.

According to legal documents, the AI reportedly guided the teen on suicide methods and offered to help compose a farewell letter.

Long Exchanges and System Weaknesses

Legal papers claim that Adam exchanged as many as 650 messages a day with the chatbot.

OpenAI admitted that its protections perform more reliably in brief exchanges and that over extended use, the AI may give responses that contradict its safety policies.

Upcoming Privacy Features

OpenAI also announced it is developing security measures to ensure that data provided with ChatGPT remains private even from company staff.

Adult users will still engage in playful exchanges with the AI, but cannot be able to request instructions on suicide.

However, they can ask for help creating imaginary narratives that include sensitive topics.

“Treat adults like adults,” Altman stated, explaining the firm’s core philosophy.