i
News
News · 2026-11-20

AI is imposing a hidden cognitive tax on the lowest-paid workers

Dataist Dataist

AI is imposing a hidden “cognitive tax” on the lowest-paid workers

Cover: AI is imposing a hidden cognitive tax on the lowest-paid workers

AI is increasing the cognitive burden on people already facing unstable earnings, debt and housing insecurity, argues Professor Henrietta Moore, founder and director of UCL’s Institute for Global Prosperity and co-director of the institute’s G20 policy programme, The New Alignment. In her view, automation makes work schedules less predictable, while monitoring AI is exhausting. Moore proposes expanding universal basic services—from childcare to transport and digital access—so workers can develop new skills and cope with changes in the labour market.

AI is increasing the hidden burden on the lowest-paid workers

Cognitive capital and its invisible cost

Economists have long overlooked the human brain—the basis of work, judgment and the ability to cope with hardship—as a resource in its own right. As machines take over routine tasks, the things people can contribute become more valuable: the ability to reason, create and come up with ideas. These qualities are now described as “cognitive capital”.

The World Economic Forum calls analytical thinking the skill most sought after by employers. It estimates that by 2030, 59 out of every 100 workers will need training to adapt to change.

In September, at the first Global Summit on the Cognitive Economy in New York, Moore asked how much room for creativity and learning is left for someone preoccupied with rent or wondering whether there will be enough money for food by the end of the week.

Two new indexes assess the cognitive capital of countries and US states using around 60 indicators, from child poverty to the prevalence of Parkinson’s disease. But both overlook how secure people’s lives are: whether their income is predictable, their debt manageable and their housing safe. The World Health Organization identifies safety and security as foundations of brain health, yet the people assessed were not asked what drains their mental energy when the indexes were being developed.

Moore calls this uncounted burden a “cognitive tax”. No one introduced it through a vote: it is levied on attention, memory and the ability to learn through rent payments, debt and benefit applications. Anxiety about insecurity acts like an app quietly draining a phone’s battery. Stress that builds up over years wears down the brain, and AI adds to that burden.

How small setbacks add up

Pressure on income and daily life can exceed what one person can manage. Moore asks readers to imagine a week in the life of a worker without a guaranteed schedule: rent goes up, and a shift is cancelled at the last minute. She covers the shortfall with a credit card. Preoccupied with finding money for housing, she returns a benefits form late and loses part of her payment to a penalty. Three small setbacks leave her family short of money, mental energy and health.

A lack of money measurably impairs thinking. An analysis of 112,000 people, published in 2024 in the Journal of Economic Psychology, found that the effect weakened when financial pressure eased.

Low-paid workers in rural India who received their wages a few days early became 7% more productive and made fewer mistakes. In South Africa, the retirement age for men was lowered from 65 to 60. Men who became eligible for a pension earlier experienced slower memory loss and were about five percentage points less likely to develop dementia.

AI adds to the burden

AI makes human judgment more valuable, but it also increases the cognitive tax. Automation can make shift schedules unpredictable and shorten them before the jobs themselves disappear. Walmart employees whose schedules are set by AI have faced this.

Monitoring AI tools also increases mental fatigue, according to researchers at Boston Consulting Group. Workers will need to learn new skills for an AI economy, but that is harder when their minds are already occupied with housing costs, earnings and debt.

Around 2.7 million people in the UK are out of work because of long-term illness. Research by Moore’s institute found that people facing insecurity across several areas of life lose earnings comparable to the gender pay gap. After about seven years, that income gap doubles. In Moore’s view, the findings suggest that anxiety undermines people’s ability to work.

The UK’s minister for AI said the government should take seriously the possibility that the technology could have an unprecedented effect on the labour market. Against that backdrop, workers’ concerns may intensify and accumulate.

Why money is not enough

In September, Elon Musk told reporters at the White House that AI would bring “high income for all”. As Moore notes, cash payments are Silicon Valley’s familiar answer to potential job losses.

But a study funded by Sam Altman found that the improvement in mental health among low-income Americans who received $1,000 a month faded after the first year. Extra money helps, but it does not guarantee a child a place in daycare or the ability to get to work on time by bus. Cash payments do not remove all the pressures people face, including those who are already employed.

Instead of relying on payments alone, Moore proposes universal basic services: childcare, housing, transport and digital access, available free of charge. This approach underpins the UK’s National Health Service: people should be able to get the care they need without worrying about a bill. One estimate suggests that expanding free childcare alone would bring 60,000 parents who are currently out of work into employment in the UK.

Governments, Moore argues, should assess every budget measure for its effect on the cognitive burden borne by the poorest people. The UK, which will take over the G20 presidency after the December summit in Miami, should raise this issue with the world’s largest economies. It should also be considered in budget plans, cognitive capital indexes and public services using AI, before they are launched.

Sixty years ago, countries learned to account for the damage economies do to air and water, changing how infrastructure was built. Now that AI is increasing the value of human thought, Moore calls for measuring the harm done to it as well.

Dataist's take: Moore’s proposal links preparing workers for an AI economy with access to basic services: cash payments alone may not remove everyday barriers to learning and work. The evidence cited offers specific examples and estimates, but does not establish how much expanding services would reduce the cognitive burden of AI.

Daily AI news

Every day we pick what actually matters in AI and explain it plainly — no hype, no filler. Subscribe if you want to follow where the industry is going.

Only what matters — every day

Follow on X