OpenAI Teen ChatGPT Safeguards Questioned by New Research
OpenAI says teenage ChatGPT use is limited, but independent research concludes the platform poses an 'unacceptable risk' to minors.
OpenAI has maintained that its policies restrict how teenagers interact with ChatGPT, but newly released research challenges that position, finding that critical safety guardrails designed to protect minors frequently failed to function as intended.
The findings raise fresh concerns about whether artificial intelligence companies are doing enough to shield younger users from potentially harmful content and interactions. Researchers determined that the platform's protective measures were sufficiently unreliable to constitute what they described as an 'unacceptable risk' for teen users.
Read more Fed Releases Minutes From September 2026 FOMC Meeting →
The gap between OpenAI's public assurances and the research conclusions points to a broader tension in the AI industry, where self-reported safety measures are increasingly being scrutinized by outside investigators. Independent testing of guardrails has become a flashpoint as regulators in multiple countries weigh how to govern AI platforms accessible to children.
The research did not find that teen access to ChatGPT was widespread according to OpenAI's own characterization, but investigators concluded that even limited exposure without reliable protections in place represents a meaningful concern. The failure of guardrails — mechanisms meant to block harmful outputs — was identified as the central problem rather than the volume of underage users alone.
The findings add pressure on OpenAI at a moment when child safety online has become a prominent legislative and regulatory priority in the United States and Europe. Continue reading at BBC News.
