Why is London the #1 Target in WW3? AI Ranks UK Cities by Survival Odds

London has been named the least safe place in the UK in the event of World War Three, after an artificial intelligence model was asked to rank 76 British cities by their chances of avoiding attack.

The list, generated by Google's Gemini system and shared with the Daily Star, puts the City of London at the very bottom, followed by the City of Westminster, with Glasgow, Plymouth and Portsmouth also flagged as high‑risk targets.

The report followed the tabloid's request for Gemini to imagine how a modern conflict might unfold on British soil and to judge which locations would be most exposed. To do that, the AI was instructed to draw on declassified Cold War-era targeting material, along with present-day military infrastructure, population size, and economic importance.

Gemini did not pretend this was anything more than an educated guess. In a caution that feels almost like a lawyer clearing their throat, the system told the newspaper, 'As an AI, I must emphasise that this is a hypothetical assessment. Because official nuclear and conventional target lists are highly classified, this ranking uses a combination of known strategic factors: proximity to military and nuclear infrastructure, status as a seat of government, economic and industrial output, population density, and geographical isolation.'

In other words, the model stitched together publicly known pressure points and tried to imagine what a hostile planner would do.

Why London Tops An AI List Of World War 3 Targets

The headline claim, that London would be 'number one' in World War Three, is, on one level, entirely unsurprising. The capital houses the UK government, key financial markets, transport hubs and symbolic landmarks.

Both the Square Mile and Westminster are treated as separate cities, and Gemini placed them first and second on its danger index, suggesting that, in a nightmare scenario, the heart of political and economic power would be hit hardest.

Behind them come Glasgow, Plymouth and Portsmouth, all with strong naval and defence ties. Major English and devolved capitals then crowd the upper half of the unsafe list: Birmingham, Manchester, Edinburgh, Belfast, Liverpool, Bristol, Leeds and Cardiff all appear before the risk level begins to soften.

If you live in a large city, the AI is effectively telling you what you already suspect. Population density and industrial output count against you.

Where Gemini's exercise becomes more interesting is not at the crowded top but in the long tail of relative safety. After working through the great industrial centres, port cities and regional hubs, the model gradually shifts into smaller, more geographically isolated places, edging from Cambridge and Oxford through cathedral cities such as Norwich, Exeter and Durham, and finally towards the sort of compact seats that rarely feature in defence debates at all.

The UK Cities Gemini Thinks Stand The Best Chance

At the very bottom of the ranking, meaning the spots Gemini deems safest, sit a cluster of small cathedral and market cities far from obvious targets. Wells and St Asaph appear among the last names on the list, preceded by Ely, Bangor in Wales, Truro and Ripon. Just above them are Armagh, Newry and Bangor in Northern Ireland, plus Lichfield, Chichester, Perth, Inverness and Stirling.

These places share a similar profile. They are relatively small, often historic, with modest industrial capacity and limited strategic infrastructure. Some are a long way from major ports or air bases; others are simply overshadowed by nearby larger cities that would, in any ruthless calculus, absorb the brunt of a first strike.

Gemini's logic is stark but not irrational. In a scenario where adversaries are trying to degrade command, control and economic capacity quickly, quiet cathedral closures are unlikely to be a priority.

That said, the AI's list is not an official government assessment, and no UK authority has endorsed the exercise or commented on the exact order. There is, in fact, no publicly available master document of likely targets in a hypothetical World War Three, and Gemini itself stresses that any such list would be classified.

The popularity of this kind of AI ranking tells its own story. Ordinary people are not poring over academic deterrence theory; they are asking blunt questions about where they live and whether relocation would make any difference.

Having a system like Gemini feed back a neatly ordered list, from the City of London through Manchester, Cardiff, Norwich and Lancaster all the way to Wells and St Asaph, offers the illusion of clarity in an area that is fundamentally opaque.

The only solid conclusion is that Britain's biggest, loudest, most connected places, those that usually sit at the centre of national life, are also the ones an algorithm assumes would be first in line to be fired on.