Oxford researchers urge EU to demand happiness dividends from subsidized AI firms
Academics argue public subsidies for artificial intelligence should yield wellbeing gains, not just corporate profits, according to a new paper.
Key points
- Oxford Internet Institute researchers propose happiness as a third goal for AI policy.
- Public subsidies include tax breaks, land access, and data used for model training.
- Companies receiving state support should be held accountable for delivering wellbeing benefits.
Researchers at the Oxford Internet Institute have published a paper arguing that European Union policymakers should require artificial intelligence companies to deliver measurable improvements in public happiness in exchange for government subsidies. The study, led by Professor Sandra Wachter, suggests that current frameworks focus too heavily on economic growth and risk mitigation while ignoring social benefits.
The paper, titled "Optimising for happiness in EU technology policy to repay public subsidies for AI," was published in Royal Society Open Science. It contends that citizens indirectly fund AI development through various means, making them stakeholders in the outcomes. The authors argue that simply reducing harm is insufficient; technology must actively contribute to social wellbeing to justify public support.
The Nature of Public Subsidies
According to the researchers, the public backs AI development through direct funding, tax breaks, and access to critical resources. These resources include land, water, energy, infrastructure, and data. The OII team notes that most people have no say in how their data is used. Creative works, social media posts, and user interactions are frequently swept up to train frontier models, a practice that has spurred a stream of lawsuits.
The authors emphasize that citizens are inadvertently helping to create the next generation of tech billionaires. They argue that this dynamic requires a shift in policy. Instead of accepting promises of economic growth, the public should expect tangible returns in the form of improved wellbeing. This approach would align corporate incentives with broader societal goals rather than just corporate profits.
Defining Happiness in AI Policy
The OII team proposes happiness as a third goal for AI heavyweights, alongside economic growth and risk mitigation. The researchers point to relatedness, competence, and autonomy as key components of happiness. They warn that some uses of AI could undermine these elements. For instance, chatbots that mimic human emotion may erode the value people place on real human connection.
Heavy reliance on AI could also dull skills and make it harder for people to make their own choices. Additionally, datacenters threaten land and water resources, leaving fewer trees and green spaces for people to connect with nature. The researchers state that making technology less harmful is not the same as making it socially beneficial. Therefore, AI policy must explicitly address how technologies impact these fundamental aspects of human life.
What should you do?
Policy observers and technology professionals should monitor how EU regulators incorporate these recommendations into upcoming legislation. Organizations receiving public funding should review their impact assessments to ensure they account for social wellbeing metrics. Stakeholders should watch for new accountability requirements that may affect how AI benefits are measured and reported to the public.
What to do and how to stay safe: Oxford Internet Institute
- Review public funding agreements for clauses related to social impact or wellbeing metrics.
- Monitor regulatory updates in the EU regarding AI subsidies and accountability standards.
- Assess internal AI usage for potential impacts on user autonomy and skill development.
- Track legal developments involving data usage for model training and related lawsuits.
Step-by-step guide: Prompt Injection Warning Signs: How to Spot AI Manipulation
General security guidance from the Threat Times newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
What is the main argument of the Oxford researchers?
They argue that public subsidies for AI should be repaid in wellbeing and happiness, not just economic growth.
How do citizens indirectly fund AI development?
Through funding, tax breaks, and access to land, water, energy, infrastructure, and data used for training models.
What are the three components of happiness cited by the researchers?
The researchers point to relatedness, competence, and autonomy as key factors that AI could undermine.
