Regulators Issue Final Warning to AI Firms Over Age Assurance Compliance

The recent coordinated warning from international regulators marks a definitive end to the 'grace period' for artificial intelligence firms regarding child safety and data privacy. For the past two years, developers of Large Language Models (LLMs) and generative AI applications have operated in a regulatory gray area, often relying on simple self-declaration checkboxes to verify user age. However, as of March 2026, regulators including the UK’s Information Commissioner’s Office (ICO) and Ofcom have signaled that such 'soft' checks are no longer sufficient to meet the stringent requirements of the Online Safety Act and the Age Appropriate Design Code.

This regulatory pivot is driven by increasing evidence that minors are bypassing existing safeguards to access age-inappropriate content, ranging from harmful medical advice to sexually explicit AI-generated imagery. The warning specifically targets 'high-risk' AI systems that are 'highly likely' to be accessed by children, a definition that now encompasses most consumer-facing chatbots and image generators. Regulators are demanding that firms implement 'age assurance' technologies—such as facial age estimation or third-party identity verification—that provide a higher degree of certainty than traditional methods while maintaining strict user privacy.

“

Under the UK’s Online Safety Act, Ofcom has the power to levy fines of up to £18 million or 10% of a company’s global annual turnover, whichever is higher.

The implications for the AI sector are profound. Companies found in breach of these requirements face catastrophic financial penalties. Under the UK’s Online Safety Act, Ofcom has the power to levy fines of up to £18 million or 10% of a company’s global annual turnover, whichever is higher. For tech giants with multi-billion dollar revenues, these fines represent a material risk to their bottom line. Furthermore, regulators have hinted at the possibility of 'stop-processing' orders, which would effectively force an AI service to go offline in specific jurisdictions until compliance is proven. This 'nuclear option' underscores the seriousness with which authorities are now viewing the intersection of AI and child safety.