Researchers at the Oxford Internet Institute, led by Professor Sandra Wachter, are urging EU policymakers to demand a "happiness dividend" from AI companies that benefit from public money. Their paper, published in Royal Society Open Science, argues that citizens already support AI through subsidies, tax breaks, and access to land, water, energy, and data — yet get little say in what they receive in return.
The team proposes happiness as a third goal for AI development, alongside economic growth and risk mitigation. They identify relatedness, competence, and autonomy as key components of well-being that AI could undermine: chatbots may erode real human connection, heavy reliance could dull skills, and AI influence could constrain personal choice. Datacenters also consume land and water, reducing access to nature.
Under their approach, companies receiving public support would predict how their technologies will improve happiness and be held accountable for delivering those benefits. The paper notes the EU has allocated €20 billion for AI gigafactories and €700 million through Horizon research projects into generative AI and supercomputing. The researchers stress that making technology less harmful is not the same as making it socially beneficial, though the report's own authors concede that profit motives may ultimately outweigh these concerns.