London · Population 8,799,800 · GVA £503,000m
London's extraordinary economic diversity is its ultimate shield — professional and scientific services at 13.2%, information and communications at 8.4%, financial services at 7.8%, and a cultural and creative sector without parallel in Europe. No single sector dominates, and the highest-risk sectors (admin at 10.4%, retail at 7.2%) are proportionally lower than in any other city in the dataset. London will not just survive AI disruption — significant portions of its economy will thrive on it, as the city is home to both the companies developing AI and the professional services firms advising on its deployment. But the aggregate statistics mask an enormous human-scale problem: 10.4% admin support in a city of 8.8 million means roughly 400,000 people in roles that AI can substantially transform. In absolute numbers, London faces the largest workforce transition challenge in the country, concentrated in outer boroughs where the economic profile looks more like a mid-table city than the capital.
London will be fine. London is always fine. The Square Mile will adapt, Canary Wharf will pivot, Shoreditch will produce AI startups that get acquired for billions, and the creative industries will find ways to make AI look edgy and sell it back to everyone else. Professional services at 13.2%, tech at 8.4% — the core London economy is literally the people building and advising on the technology that's displacing everyone else. But here's what doesn't appear in the 'London is resilient' narrative: the 400,000 admin workers in Croydon, Barking, Hounslow, and Newham whose jobs are the same admin jobs being automated in Sunderland, just with a higher rent burden. London's average conceals its extremes. Tower Hamlets has a financial services economy and Newham has a logistics economy, and they share a borough boundary and approximately nothing else. The City will throw a conference about 'responsible AI' at the Barbican while the actual displacement happens in outer London offices that nobody at the conference has visited. London doesn't have an AI problem. It has 32 different AI problems, one for each borough, and the ability to paper over every single one with aggregate statistics.
London's challenge isn't resilience — it's inequality. The tech, professional services, and creative sectors will thrive, but the admin, retail, and logistics workers face displacement at unprecedented scale in absolute numbers. The Mayor's office needs a city-wide retraining entitlement that targets the outer boroughs where the employment profile resembles Birmingham more than the City of London. Borough-level transition programmes should be funded proportionally to displacement risk, not population. The tech sector should be leveraged as a source of transition training: every major AI company operating in London should contribute to a workforce transition fund. London's cultural and creative sectors need protection from the commercial pressures that push them out of the city — affordable workspace guarantees, business rate relief, and recognition that the creative economy is an employment strategy, not a lifestyle amenity.
They'll announce an 'AI Strategy for London,' host a summit at the Barbican, produce a glossy PDF with a foreword by the Mayor, and distribute it at a reception where the canapés cost more than a week's wages for the admin workers whose futures are under discussion. Each borough will produce its own AI readiness assessment, none of which will be coordinated with each other. Canary Wharf will pivot to 'AI-friendly workspace' by adding some power sockets and calling it innovation. Shoreditch will produce a hundred AI startups, ninety of which will fail, nine of which will be acquired, and one of which will be worth billions and employ fifty people in London and two thousand in India. The actual admin and retail workers — the ones in Ilford and Barking and Croydon — will find out about the future of work from a government website and a JobCentre appointment that offers them a link to an online course. The congestion charge will go up. The tube will get more expensive. And London will continue its ancient tradition of being the best place in the world to be rich and an increasingly difficult place to be anything else.
Employment share by SIC sector, with automation risk weight and contribution to overall score. Sectors with higher risk weights contribute more to the vulnerability score.
| Sector | Employment % | Risk Weight | Risk Tier | Contribution |
|---|---|---|---|---|
| Professional, Scientific & Technical | 13.2% | 0.3 | low | 3.96 |
| Human Health & Social Work | 10.8% | 0.18 | low | 1.94 |
| Administrative & Support Services | 10.4% | 0.85 | high | 8.84 |
| Education | 8.8% | 0.15 | low | 1.32 |
| Information & Communication | 8.4% | 0.5 | medium | 4.2 |
| Financial & Insurance Services | 7.8% | 0.75 | high | 5.85 |
| Accommodation & Food Services | 7.6% | 0.48 | medium | 3.65 |
| Retail | 7.2% | 0.8 | high | 5.76 |
| Public Administration & Defence | 5.2% | 0.22 | low | 1.14 |
| Transport & Storage | 4.8% | 0.78 | high | 3.74 |
| Arts, Entertainment & Recreation | 4.1% | 0.2 | low | 0.82 |
| Construction | 3.8% | 0.28 | low | 1.06 |
| Wholesale | 3.2% | 0.55 | medium | 1.76 |
| Manufacturing | 2.4% | 0.82 | high | 1.97 |
| Real Estate | 2.1% | 0.4 | medium | 0.84 |
| Agriculture, Forestry & Fishing | 0.1% | 0.25 | low | 0.03 |
The vulnerability score is a weighted average of London's sector employment shares. Each sector carries an automation risk weight (0.0–1.0) derived from Frey & Osborne's occupational automation probabilities, mapped to SIC sectors via ONS correspondence tables. The weighted average is then normalised to a 0–100 scale. A score of 62.3 means London's workforce is significantly concentrated in automatable sectors compared to other United Kingdom cities.