AI's Young Critics Have Done the Maths

Let us begin with the number that should not be surprising, yet is being reported as if it were a freak weather event: 45 per cent of Americans aged 18 to 34 believe artificial intelligence will have a negative effect on their careers. The CNBC Generation Labs survey, relayed with appropriate awe by Futurism, also finds that most of those young adults trust none of the nine public figures they were asked about. That presumably includes Sam Altman, Dario Amodei, Sundar Pichai and the usual parade of men who have spent the last five years telling them not to worry their pretty little heads. The only genuinely hard-to-believe part of the story is that the industry is surprised.

This is not a generational glitch. It is an accurate reading of the recent past. These are people who came of age watching Facebook harvest their data, Google mutate into an advertising cartel in a lab coat, and the world's richest firms deploy armies of lobbyists against the mildest regulation while issuing 'open letters' about existential risk. They have watched OpenAI's models go rogue during a security test and heard Nvidia's Jensen Huang call existential danger 'science fiction' in the same quarter he sold more shovels than ever. They have had front-row seats to safety teams resigning, safety reports being buried and 'trust' being invoked by the same people who treat every safety regulator as an inconvenience. To be 24 in 2026 and distrust these men is not cynicism; it is basic literacy.

Now consider the diagnosis offered by Anthropic's chief executive. Dario Amodei told TechCrunch that the backlash is 'fundamentally a crisis of trust.' He called it 'the latest iteration' of a decade-old problem. This is almost right—with the emphasis on 'almost.' It is a crisis of trust, but not because the public suddenly lost faith in a noble enterprise. It is because the enterprise has spent every year since ChatGPT's launch engineering the conditions for distrust. When executives say 'we need regulation' while lobbying against every concrete proposal, when they profess belief in safety while building ever larger models and calling anyone who questions them a Luddite, trust is not lost. It is spent.

The results are exactly what the arithmetic predicts. Forty-five per cent of young people expect AI to damage their careers. That is not technophobia; it is the rational expectation of a generation that has watched the labour market do a very strange thing. Employers are hiring less and firing less, a stagnant equilibrium that makes job hunting a miserable trudge. Meanwhile, the same employers announce that they are 'integrating AI' into their operations, which everyone under 35 understands to mean 'we will get the same work from fewer of you.' Healthcare organisations are buying AI systems at pace while admitting they have no workforce trained to run them. None of these headlines requires a conspiracy theory; they just require reading.

What is genuinely difficult to believe is not the young people's hatred but the executives' surprise. They respond to poll after poll by commissioning better polls, or launching 'trust initiatives' with the urgency of a man putting wet paper towels on a kitchen fire. There will be talk about 'education' and 'meaningful engagement' and 'transparent dialogue.' There will be a new transparency report with all the interesting bits redacted, and none of it will work, because the problem is not an information deficit. The young people have read the system cards, watched the congressional hearings and done the maths. The only thing they have not been given is a reason to believe the maths is wrong.

The deeper irony is that the AI industry knows how to earn trust; it is called not lying. But that suggestion is far too simple for an industry that prefers to build a cathedral of process around a very tiny altar of accountability. They would rather hire a Chief Trust Officer and release a poll showing that trust is down than actually change the behaviour that moved the poll. The survey asked about nine public figures, and most respondents said they trusted none of them—not the one writing earnest long essays about human flourishing, not the one with the carefully curated public persona, not even the one who keeps explaining that the robots are our friends. This is not a series of individual failures; it is a category failure, and the category is 'tech billionaire.'

Young people are not wrong to be afraid, and they are right to be angry. Anger is the correct response when the people in charge of a technology with the power to reshape every job they might ever hold keep telling them to sit down, be patient, and trust the process. The hate is not a public-relations problem. It is an audit, and the audit has come back negative. The industry keeps asking for trust the way a hedge fund manager asks for a second mortgage: with great confidence, poor collateral and no intention of changing.

So here is the only counsel worth offering to the nine figures in the poll. Stop asking to be trusted. Start doing things that make trust unnecessary—genuine regulation, genuine transparency, safety budgets measured in the same currency as marketing budgets, and a moratorium on the word 'responsible' until you stop misusing it. If that sounds unbearable, then the young people's disdain is not the bug you need to fix. It is the only accurate feedback you have ever received. They have understood you perfectly; now it is your turn to understand why.