The UK government has pushed back its plans to regulate artificial intelligence by at least a year, opting instead to prepare a more wide-ranging bill covering both safety and copyright issues.

Technology Secretary Peter Kyle intends to introduce the legislation in the next parliamentary session, but it will not be ready before the next King’s Speech — now expected in May 2026 — raising concerns about delays in tackling the risks posed by powerful AI systems.

The Labour government had initially planned to move quickly with a narrow bill focused on large language models like ChatGPT, which would have required companies to submit their models for safety testing by the UK’s AI Safety Institute. That proposal has now been shelved, partly to allow time for alignment with Donald Trump’s administration in the U.S.

Ministers say the upcoming bill will also address mounting pressure over the use of copyrighted material in AI training. The government is currently facing strong resistance from the creative sector over its separate Data Bill, which would allow AI companies to use copyrighted works unless creators opt out. Artists like Elton John, Paul McCartney, and Kate Bush have publicly opposed the change.

Peers in the House of Lords recently backed an amendment requiring AI firms to disclose whether they use copyrighted content in training — a move the government has resisted. Kyle has since committed to launching a cross-party working group on AI and copyright and said the full issue will be dealt with in the new bill.

While critics say the delay leaves the UK behind in regulating a fast-moving technology, Kyle argues that a more comprehensive approach will allow space to “find solutions” in dialogue with both tech companies and rights holders.

Civil society groups and public surveys show widespread concern: an Ada Lovelace Institute–Alan Turing Institute poll found that 88% of Britons believe the government should have the power to block AI products that pose serious risks. Over three-quarters want regulators — not private firms — overseeing AI safety.

Scott Singer, a senior fellow at the Carnegie Endowment, said the UK is walking a fine line: “Like the U.S., Britain is trying to avoid aggressive rules that could stifle innovation, while still protecting consumers. That’s the balancing act here.”

Copied