Daron Acemoglu, an economist and Institute Professor at the Massachusetts Institute of Technology (MIT), has warned that the current trajectory of artificial intelligence (AI) development risks deepening inequality and accelerating job losses — unless industry pivots toward tools that complement, rather than replace, human workers. Writing in The Humanist Review of AI, Acemoglu argues that the policy conversation must shift from speculative fears about Artificial General Intelligence (AGI) to concrete, pro-worker frameworks that guide AI toward human flourishing.

The Core Argument

Acemoglu contends that the framing of AI discourse is itself part of the problem. Debates dominated by either utopian AGI promises or dystopian civilisation-ending fears distract from the immediate, tractable issue: who benefits from AI as it is being deployed today. “The most important tool for steering AI onto a better path is not a silver bullet policy,” he wrote, calling instead for a fundamental change of perspective — one that prioritises complementing human skills over automating them away.

Acemoglu defines Artificial General Intelligence (AGI) as a hypothetical AI system capable of matching or surpassing human cognitive abilities across any intellectual task or domain, stressing that current concerns about it are largely speculative and distract from present-day harms.

What Pro-Worker AI Looks Like

The paper outlines a four-point policy agenda to redirect AI development. According to the paper, governments and institutions should: “Create corporate demand for pro-worker tools; generate demonstration effects that prove the approach works; build a market environment where new ideas and business models can flourish; and invest in the infrastructure for pro-worker AI.”

Crucially, Acemoglu argues these are not simply regulatory prescriptions — they are structural interventions designed to shift what the market incentivises. The goal is to make it commercially viable, not just ethically desirable, to build AI that augments workers rather than eliminates them.

Why Big Tech Is Part of the Problem

The paper squarely blames the dominant business models of large technology companies for the current impasse. Acemoglu argues that firms focused on selling software to corporations and maximising revenues from digital advertising have had little commercial incentive to develop pro-worker tools. Compounding this, the AI industry has become heavily concentrated, leaving scant room for new entrants to experiment with alternative business models.

A further structural trap, which the paper describes as a ‘nuts and bolts’ problem, reinforces the status quo: businesses assume they will only be offered automation tools, so they plan accordingly — and tech companies, reading that demand signal, invest further in automation. “Tech companies think that businesses will continue to predominantly demand automation tools and there wouldn’t be a large market for pro-worker AI, and they invest and develop their models accordingly,” the paper noted.

Public Anxiety and Productivity Doubts

Acemoglu cites data showing that nearly 52 per cent of US citizens are worried about how AI will impact their jobs. Beyond employment, broader anxieties include AI’s potential to pollute the informational ecosystem, undermine democratic institutions, and — in more extreme forecasts — pose existential risks to civilisation.

Yet, in a finding that challenges both AI optimists and pessimists, Acemoglu notes that the promised productivity revolution has so far failed to materialise in measurable terms. “Several studies and reports have failed to find much in the way of productivity gains from adoption of AI by most companies,” he observed — suggesting the economic case for a wholesale automation push is weaker than its proponents claim.

Broader Implications

This comes amid a period of intense global debate on AI governance, with regulators in the European Union, the United States, and India all weighing frameworks to manage AI’s social and economic consequences. Acemoglu’s paper adds an economist’s rigour to a debate often dominated by technologists, arguing that getting the incentive structures right — not just the regulations — will determine whether AI becomes a tool of broad prosperity or concentrated wealth. The direction that policymakers, investors, and technology leaders choose in the near term, the paper implies, may be difficult to reverse.