British regulators have indicated they are preparing a more assertive approach to artificial intelligence governance, amid growing calls from industry leaders for clearer rules that balance innovation with public protection.
The Financial Conduct Authority and the Competition and Markets Authority have both published discussion papers in recent weeks outlining their intent to apply existing regulatory frameworks to AI-driven products and services. The move follows months of consultation with technology firms, consumer groups, and academic researchers.
“What we are seeing is a shift from voluntary guidance towards a more structured expectation of accountability,” said Dr Helena Markham, senior research fellow at the Centre for Data Ethics and Innovation. “Firms can no longer treat AI governance as an optional bolt-on. It is becoming central to how regulators assess operational resilience.”
The UK has so far resisted following the European Union’s comprehensive AI Act model, instead favouring a principles-based framework that empowers individual sector regulators. However, industry insiders suggest the government may introduce targeted legislation before the end of the parliamentary session, particularly around high-risk applications in recruitment, credit scoring, and biometric surveillance.
A survey of 500 UK tech firms conducted by TechUK found that 68 per cent of respondents support clearer statutory obligations, provided they are proportionate and do not disadvantage smaller enterprises. The same survey found that 41 per cent of firms have already appointed dedicated AI ethics officers in anticipation of tighter oversight.
“Businesses want regulatory certainty,” said James Coulson, policy director at the British Technology Association. “Investment decisions are being delayed because firms do not know what the compliance landscape will look like in three years. A clear, pragmatic framework would unlock capital, not constrain it.”
The government’s AI Safety Institute continues to expand its testing capabilities, recently opening a new laboratory in Manchester focused on evaluating the robustness of large language models deployed in enterprise settings.