The UK government is preparing to expand online safety enforcement to AI chatbots after public pressure and regulatory gaps exposed by recent incidents.
According to The Guardian, ministers plan to close a loophole in the Online Safety Act so chatbot providers can be held directly accountable when their systems generate harmful illegal content, including material that may encourage self-harm or abuse.
What is changing
The proposed update would require AI chatbot providers to comply with illegal-content duties under the Online Safety Act, similar to obligations already applied to major online platforms.
If companies fail to comply, potential penalties include:
- fines of up to 10% of global revenue
- legal action that could lead to service blocking in the UK
Why this matters
Regulators have acknowledged current law does not fully cover all chatbot-generated outputs unless specific conditions are met. In practice, that has left a policy blind spot while usage among children has accelerated.
The government’s direction signals a broader principle: AI interfaces that shape user behaviour should face the same safety expectations as other large digital products.
Industry implications
For AI product teams, this is a strong warning to move from reactive moderation to proactive safety architecture:
- age-aware safeguards
- harm prevention by design
- clearer escalation and auditability
Source: https://www.theguardian.com/technology/2026/feb/15/ai-chatbots-children-risk-fines-uk-ban
