UK AI Update May 2026: FCA Flags Chatbot Risks, DeepSeek Banned on Government Devices

May 2026 has been a busy month for UK AI regulation, with the Financial Conduct Authority raising fresh concerns about AI-powered personal finance tools, and the UK government quietly restricting access to DeepSeek’s R1 model on public sector devices. Here’s what’s happened and what it means for anyone using AI in the UK.

For a country that has positioned itself as an open, “pro-innovation” home for artificial intelligence, the UK is navigating some genuinely difficult questions. The gap between AI’s capabilities and the regulations designed to govern it is widening — and authorities are scrambling to keep pace.

FCA Flags AI Finance Chatbots as a Regulatory Gap

The Financial Conduct Authority’s latest perimeter report has drawn attention to the rapid rise of general-purpose AI tools acting as personal finance advisers. Tools like AI chatbots that tell users which savings account to open, whether to remortgage, or how to invest a lump sum are growing in popularity — but they currently sit outside the FCA’s regulatory perimeter.

The FCA warned that “current perimeter boundaries may no longer be fit for purpose” if these unregulated services begin to cause consumer harm. In plain English: if an AI tells you to invest in something and it goes wrong, there is currently no regulatory protection for you in the way there would be if a human financial adviser gave the same advice.

This is a significant issue for anyone using AI tools to manage money. The FCA has not yet confirmed what regulatory changes are coming, but the language of the perimeter report suggests action is likely before the end of 2026.

DeepSeek R1 Restricted on UK Government Devices

The UK government has taken steps to review and restrict use of DeepSeek’s R1 model on government and public sector devices, following concerns about data practices and national security implications. DeepSeek, a Chinese AI lab that released R1 to widespread acclaim in early 2026 for its low cost and strong performance, has faced scrutiny in several Western countries.

The concern is straightforward: any data processed by DeepSeek’s models could theoretically be accessible to the Chinese government under Chinese national security law. For sensitive public sector work, that risk is considered unacceptable. The UK’s decision follows similar moves in the United States and parts of the European Union.

For private users and businesses, there is no legal restriction on using DeepSeek in the UK. But the government’s position serves as a clear signal about how AI tools developed in authoritarian states are likely to be treated in policy discussions going forward.

Microsoft and OpenAI Restructure Their Partnership

In a significant development for the global AI industry, Microsoft and OpenAI have restructured their partnership to a non-exclusive arrangement. This means Microsoft is no longer OpenAI’s sole commercial partner for cloud deployment, and OpenAI can now work with other cloud providers — including Google and Amazon Web Services.

The shift opens up greater commercial competition in the cloud AI market. For UK businesses using OpenAI’s GPT models via Azure, this change may eventually result in better pricing as competition increases. For the broader market, it signals that the OpenAI-Microsoft relationship — which had defined the AI landscape for two years — is evolving into something more arm’s-length.

House of Commons Committee Launches AI at Work Inquiry

The House of Commons Business and Trade Committee has launched a formal inquiry to examine both the opportunities and risks of AI adoption in UK workplaces. The inquiry will assess whether existing worker protection laws remain fit for purpose as generative and agentic AI accelerates across industries.

Key questions include: Can an employer use AI to monitor workers without consent? Who is liable if an AI system makes a hiring or firing decision that discriminates? Do current employment contracts cover AI-assisted work arrangements?

The inquiry is expected to run through summer 2026, with recommendations likely in autumn. Employment lawyers and HR professionals should watch this closely — the outcomes could require significant updates to employment contracts, data protection policies and AI use agreements.

What the UK’s AI Stance Means for Businesses

Taken together, May 2026’s AI regulatory activity paints a consistent picture: the UK wants to lead on AI adoption, but is increasingly aware that it needs guardrails. The FCA’s chatbot warning, the DeepSeek restrictions, and the parliamentary inquiry all point toward a more active regulatory posture in the second half of 2026.

For UK businesses using AI: document your AI use cases now, understand which tools process personal or sensitive data, and be prepared for sector-specific guidance from regulators like the FCA, ICO and CMA over the coming months.

What This Means for You

If you’re a UK consumer using AI tools for financial guidance — whether that’s asking ChatGPT where to invest or using an AI-powered comparison tool — be aware that these services currently fall outside regulatory protection. Use them for general information, but treat personalised suggestions with the same scepticism you’d apply to advice from a stranger on the internet.

If you’re a business using AI: now is the time to conduct an AI audit. Map out every tool you use, who processes the data, and whether your existing policies cover it. Getting ahead of regulation is far less painful than reacting to it.

This article is for educational purposes only and does not constitute financial or legal advice. Always seek professional advice for your specific circumstances.