A professor of economics at the University of Oxford has argued that while much attention focuses on AI’s potential to suppress wages or displace jobs, the wider societal consequences – including a fundamental shift in power – are being overlooked.
Jean-Paul Carvalho, speaking at the UK Finance Digital Innovation Summit, warned that several risks have “flown under the radar” and could destabilise democratic institutions.
“The winners of the AI revolution could gain influence through lobbying efforts. As labour is increasingly replaced by AI-generated income, taxation will constitute a smaller share of government revenue, making the AI sector more critical for state financing,” he said. “When production and innovation are no longer tied to the size and skill of the human population, why would the beneficiaries of AI fund provisions like universal basic income?”
Carvalho warned that many societal benefits currently taken for granted, such as funding for scientific research, could diminish.
Another overlooked risk involves the international balance of power. Carvalho suggested that a nation dominating AI could destabilise other economies by orchestrating labour strikes remotely.
“Imagine restructuring your entire organisation to rely on AI agents, only to find that the underlying intelligence is controlled by a foreign firm or state,” he explained. “They could deny service, throttle performance, or degrade the quality of intelligence without your knowledge.”
Carvalho noted that China and the US lead in AI investment, with China ahead in patents, academic publications, and industrial robotics.
He also cautioned that AI could erode human productivity, drawing parallels to historical examples such as the decline of indigenous skills after colonialism.
“People may invest less in developing skills if they cannot compete with AI,” he said. “Much of our expertise comes from learning from others. If individuals stop cultivating these abilities, that collective knowledge disappears.”
