Anne Gerdes: Tech Industry Hijacks AI Ethics Research Agenda

When AI ethics researchers turn to technical methods to bring ethics into AI systems, they may be solving a problem of the tech industry's choosing. In a 2022 paper in Discover Artificial Intelligence, Anne Gerdes of the University of Southern Denmark argues that the ethics research community faces two interlocking challenges that have kept it working within a framework that serves industry rather than the public.

The first challenge is direct: the tech industry has been "heavily influencing the AI ethics research agenda," shaping which questions researchers ask and what counts as acceptable answers. The second is subtler. The community's response, turning to "value-driven design methods to bring ethics to AI design," has compounded the problem. By framing research questions within a technical practice, researchers have "facilitated the technological solutionism behind the tech industry's business model" -- ethics becomes a design layer rather than a political contest over whether a system should exist at all.

The paper also addresses what Gerdes calls conflicts of interest in public policymaking on AI, suggesting industry influence extends beyond setting research priorities to shaping how governments approach regulation. Specific companies, funding arrangements, or policymaking bodies are not named, so the argument is structural rather than documented, and readers should hold the diagnosis with that caveat in mind.

Gerdes' proposed alternative is an emancipatory framework that "brings politics to design" and insists that AI should not be "treated as an inevitability." The paper is pointed about who absorbs the cost when that framing is absent: "the vulnerable groups seldom heard, despite the fact that they are the ones most negatively affected by AI initiatives," along with what Gerdes calls the "repressive power dynamics exacerbated by AI" that the research community has so far underaddressed.

The paper does not resolve the institutional question of what an emancipatory AI ethics research community looks like in practice, or who funds it independently of the industry it is meant to scrutinize. Public research funders, civil society organizations, and scholars willing to treat AI deployment as a political choice rather than a technical given are the actors best placed to take that work on.