An MP has urged the Government to use the UK’s 2027 G20 presidency to pursue an international agreement prohibiting the development of superintelligent AI. The minister’s reply pointed instead to a global code and principles for safe AI, leaving a clear gap between the call for a ban and the Government’s stated approach.
Watch Desk analysis
What happened
At Commons departmental questions on 15 October, Liberal Democrat MP Olly Glover asked AI minister Kanishka Narayan whether the Government would back the proposal. Glover said more than 85 parliamentarians had called for an international agreement, according to Resultsense. He also said AI could transform research in his Oxfordshire constituency, home to the Harwell and Culham science campuses, if its risks were taken seriously.
Narayan did not commit to a prohibition. He said the Prime Minister had put AI at the top of the G20 agenda and that the UK would seek a leading role in agreeing “a global code and principles for safe AI”. In a separate answer about regulation, he pointed to existing work on data protection, competition, equality law and online safety, while saying he would work with MPs on any further measures needed.
Why it matters
The difference is not a small drafting quibble. A prohibition would seek to stop development of superintelligent AI; a code and principles could guide how countries manage AI without ruling out that development. The exchange does not settle what the UK will take to the G20, but it shows the Government has not endorsed the specific ban Glover asked about.
For researchers, companies and the public, the eventual distinction between agreed principles and enforceable limits will matter more than a fine phrase in a summit statement. The UK’s presidency gives it a platform, not a policy commitment. Those are often mistaken for each other, particularly when international diplomacy is involved.
Our read
This is an early policy signal, not a Government plan to ban superintelligence or a settled international proposal. The useful point is the contrast: a backbench call for prohibition met a ministerial answer centred on safe-AI principles and existing laws. Readers should watch what ministers put on the G20 agenda, and whether the Government later says what those principles would actually require.
What to watch
- Whether the Government formally responds to the parliamentarians’ call.
- What AI commitments the UK proposes for its 2027 G20 presidency.
- Whether any proposed global code includes binding restrictions on development, or relies on voluntary principles.
- Whether the Government identifies additional domestic measures beyond the legal areas Narayan listed.
Discussion spark: Should the UK use its G20 presidency to seek a binding prohibition on developing superintelligent AI, or prioritise an international code that countries are more likely to accept?
Sources and evidence
- MP urges a G20 push to prohibit superintelligent AI – Resultsense (19 October 2026, 06:30 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.