TL;DR:

• A new FT-Focaldata poll of 4,000 US and UK workers shows more than 60 per cent of the highest-paid workers use AI daily compared with just 16 per cent of the lowest earners — a starkly unequal pattern of adoption.
• The data also reveals a persistent gender divide, with men significantly more likely than women to use AI tools across technology, education and retail.
• Nobel laureate Daron Acemoglu called AI “almost for sure” going to increase inequality between labour and capital; the findings also validate Rishi Sunak’s same-day BBC comments on flattening junior-role hiring in law, accountancy and creative sectors.

The Financial Times, working with research firm Focaldata, has launched a new monthly AI workforce tracker based on a poll of 4,000 US and UK workers. Its first release shows AI adoption in the workplace is heavily skewed towards the highest earners: over 60 per cent of top-paid workers use AI daily, compared with just 16 per cent of lower earners.

What the data actually shows

The pattern is striking in two directions. First, the pay gap: the strong relationship between income, education and AI use suggests AI may increase earnings inequality by boosting top-tier productivity without lifting the bottom. Second, the occupational concentration: lawyers, accountants and software developers use these tools at similar rates whether junior or senior — but much more heavily than lower-paid workers in the same industries. Corporate training is the single biggest driver of uptake in the FT data.

A persistent gender divide runs through the numbers. The FT finding aligns with data from Google’s chief economist Fabien Curto Millet showing women are 20 per cent less likely to use AI than men. Google research on structured AI training sessions for UK women over 55 showed daily usage tripled — suggesting the gap may be malleable with the right intervention.

The experts’ voices are unusually sharp. Nobel laureate Daron Acemoglu told the FT: “AI is going to increase inequality between labour and capital. That is almost for sure.” Chris Pissarides, another Nobel-winning economist at LSE, argued that as AI gets more intelligent, “your IQ matters more and more”. Oxford’s Carl Benedikt Frey noted the same pattern played out during the PC revolution but the gap eventually closed — leaving the question of how long the AI-era closure takes.

The UK connection

Two UK signals published this week sharpen the meaning. Former PM Rishi Sunak told the BBC the same morning that AI is flattening hiring for young people in law, accountancy and creative sectors — categories exactly where FT data shows the heaviest AI use. AWS UK research, released this week, found 49 per cent of UK organisations cite skills shortages as their main challenge and 24 per cent reach advanced AI use stages — consistent with a picture of concentrated AI-skill value.

OpenAI chief economist Ronni Chatterji told the FT that the heaviest AI users are not the youngest workers but those in their thirties with longer tenure — meaning AI “complements proficiency”. That flips a common assumption about generational adoption and is especially important for UK graduate-hiring discussions.

Looking forward

For UK firms, the policy question is whether structured AI training — the one intervention the FT data actually shows closing the gap — can be scaled fast enough. The Department for Work and Pensions, the Department for Education’s apprenticeship route, and the Sovereign AI Unit’s newly-backed startups all sit adjacent to this problem. The FT plans a monthly release cycle, so the tracker will be the clearest UK-focused public dataset on AI workplace inequality for the next year. Expect it to become a standard reference point for UK policy.