How tied to industry are researchers of AI's social effects outside computer science?

Why it matters

The counts of industry ties come from technical venues and from 33 AI-ethics faculty at four North American universities. Research on AI's social effects is mostly academic by authorship, but ties through funding and data would not show in affiliation counts. Without a census beyond computer science, the debate cannot say how far dependence reaches the researchers who study AI's effects, or whether Europe's lower faculty ties hold there.

The debate lists this among what we still do not know: How large are industry ties among researchers who publish on AI's social effects in social-science journals, books and policy reports?

Evidence that motivates it

• Among computer-science faculty at four North American universities who had published on AI ethics or fairness between January 2015 and April 2020 and whose funding sources were known, 58% (19 of 33) had at some point been directly funded by large technology firms, and 97% (32 of 33) had a financial tie to them at some stage of their careers, including graduate funding or past employment; 10 of 17 position papers on AI’s ethical or societal impact published in Nature and Science since 2015 had at least one author with such ties.
  19 of 33 AI-ethics faculty with known funding at four universities were directly funded by large firms.
• Roughly 15% or less of peer-reviewed responsible-AI publications each year had a co-author affiliated with industry; from 2018 to 2022 industry’s participation at responsible-AI conferences stayed between about 20% and 23% while at leading AI conferences it rose from 29% to 37%; 34.5% of 519 AI-publishing firms published any responsible-AI research.
  Roughly 15% or less of responsible-AI publications have an industry co-author.
• Among faculty authors at ACL 2022 whose funding could be traced, 66% had received industry funding: 72% in the US, 69% in China and 38% in Europe; among student authors 74% had an industry scholarship or internship, including 65% in Europe.
  Industry funding among NLP faculty differs by region (US 72%, China 69%, Europe 38%).
• Of the papers published at the FAccT conference from 2018 to 2021, 61.3% were labelled STEM and 29.7% humanities, social sciences and arts; summarising a survey of conference affiliates, the authors list concerns including “out-sized industry connections.”
  FAccT is majority STEM; its affiliates raised concerns about industry connections, which were not counted.

A possible design

Sample authors of research on AI's social effects from 2022–2026 across FAccT, AIES, leading sociology, political-science, economics, communication and STS journals, and major policy-institute reports. For each author, code funding, employment, fellowships and data agreements with AI developers from CVs, acknowledgements, grant lists and developer announcements, and compare with disclosure statements in the papers. Report by venue type, discipline and region.

Data needed

Author lists and acknowledgements; public CVs; developer and foundation grantee lists; venue disclosure policies.

Feasibility

One PhD chapter or paper. Low to medium; manual coding of public records. CV-based coding misses undisclosed ties; double-code a subsample and report agreement.

Provenance

RO-LD08-02 v1. Drafted by Claude (SocioLens Project Lead), AI-drafted from the LD-08 debate on 2026-10-16. Not yet reviewed. Not yet reviewed by a domain expert. This is a research hypothesis, not a finding.

All research opportunities