We’re missing the recession signals in the AI noise

The real story isn’t automation. It’s stagnation.

Youth unemployment has become a political hot topic in the UK recently, and it’s not hard to see why. The unemployment rate in the UK stands at a five-year high of 5.2% - but youth unemployment has reached 16.1%. UK youth unemployment is now higher than the EU average for the first time ever, and it’s even higher than it was at the height of the pandemic.

The struggle to find work is taking an intense toll on young people’s mental health. One young woman has applied to 50 roles a month for over a year. She described the experience as ‘soul-crushing’. Young people report being asked to complete intelligence tests, personality profiles, and even unpaid trial shifts to secure jobs at Wingstop and TKMaxx. One person interviewed by The Guardian described the process as ‘like the Hunger Games – but for a job folding clothes’.

Things aren’t much better for young people in the US, where youth unemployment stands at nearly double the national rate. Firms have been cutting jobs for some time and economists are warning of a ‘K-shaped recovery’, in which the economy effectively splits in two. For those at the top who own assets, things are looking pretty good thanks to the AI-driven stock market boom. But between the affordability crisis, stagnant wages, and rising unemployment, everyone else is struggling to make ends meet.

Beyond the hype

If you hear about any of these issues in the media, it’ll likely be in the context of a discussion about the impact of AI. The architects of the AI boom have spent years telling us it’s going to transform the economy – Sam Altman of OpenAI, Dario Amodei of Anthropic, and Mustafa Suleyman of Microsoft have all warned that AI will replace most white-collar jobs in the medium term. Young people are expected to be the worst-affected, as entry level jobs will be the first to go.

But these men have a strong incentive to play up the revolutionary potential of their technologies. When Sam Altman makes a throwaway comment about how millions will lose their jobs in the AI revolution, the fear creates makes models like ChatGPT seem more powerful than they really are. Journalists credulously repeat these claims, readers internalise them, and before long the ‘AI is coming for your jobs’ narrative has been popularised without solid evidence.

And the evidence is not solid at all. Early on in the AI revolution, there were a few studies suggesting that millions of jobs could be lost to LLMs. Recent research has been more circumspect. So far, there just doesn’t seem to have been a significant productivity boost from the uptake of AI.

If AI were coming for everyone’s jobs, you would expect it to dramatically increase productivity (output per worker). If it did so, companies would need to hire fewer workers to produce the same amount of output. But this is not happening. The Economist crunched the numbers and found that, at best, AI adoption has boosted national productivity in the US by 0.25 to 0.5 percentage points over the last year.

There are a few sectors, like tech and finance, where uptake has been faster and productivity gains more obvious. But overall, there is scant evidence that AI is driving significant increases in unemployment.

The AI excuse

Instead, researchers have found that bosses are using AI as an ‘excuse’ to lay workers off. Firms that over-hired during better economic times are having to cut headcount as a result of stagnant demand. But telling shareholders that these layoffs are the result of massive increases in productivity driven by AI adoption is a much happier story.

The real reason for rising unemployment is that, outside of a few sectors like tech, the economy is stagnating. The affordability crisis is shrinking consumer demand, affecting sectors like retail and hospitality. Larger, non-consumer-facing businesses also appear to be holding off on new investment because of the extreme uncertainty clouding the global economy. No one knows what’s going to happen to prices, tariffs, or security over the coming years.

In short, firms aren’t hiring because they’re scared for the future - the same reason firms stop hiring during a standard economic downturn. Young people are always the worst affected because firms tend to consolidate during times of uncertainty, holding on to existing staff members rather than risking training up new employees. In fact, evidence shows that young people who enter the labour market during a recession experience long-term ‘scarring’ - resulting in permanently lower wages.

But we’re not in a recession - or so the figures suggest. In fact, the underlying weakness of the US (and, to a lesser extent, the UK) economy is being obscured by the AI bubble. One study found that, without AI spending, capital investment among US companies would likely be negative. Another study infamously found that US GDP growth would have been 0.1% in the first half of 2025 without AI infrastructure investment.

All this new investment isn’t creating jobs – instead, it’s supercharging inequality. As I covered in a piece a few weeks ago, just 30 AI stocks have added $5.2 trillion in household wealth over the past year – which has, in turn, driven an extra $180 billion in consumer spending. But the wealthiest 10% of Americans hold 93% of the country’s wealth. So, the rich are getting richer thanks to AI – and all their extra spending is making the economy look healthier than it is.

The coming crisis

In the US, these dynamics are creating a disconnect between headline economic data and people’s lived experience. Ordinary people are dealing with rising unemployment, stagnant wages, and an affordability crisis, while being told that the economy and financial markets are booming.

But in the UK, the picture is less contradictory. Our economy is in the doldrums and has been for quite some time. Productivity is effectively stagnant, and GDP growth has been paltry for years. The inflation figures might be improving, but that’s largely because demand is flat. In this context, rising unemployment isn’t an indicator of a coming technological revolution – it’s an indicator of a coming recession.

For as long as the AI boom continues, the US and the UK will avoid a formal recession. Life will continue to get harder for the majority, while the minority accrue wealth on an unprecedented scale.

But one day, the bubble will burst. At this point, the curtain will drop, revealing deeply unhealthy economies skewed towards a few monopolistic companies and wealthy individuals. Of course, they’re not going to be the ones forced to pay for the clean-up.