Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Publishing Integrity topic
No spam. Unsubscribe anytime.
House Science subcommittee warns paper mills, AI and business models threaten the integrity of scholarly publishing
Summary
At a congressional hearing, lawmakers and witnesses warned that paper mills, predatory journals and AI-generated content are contaminating the scientific record and urged stronger oversight, more transparency and better resourcing for investigatory offices.
Chairman McCormack opened a House Science subcommittee hearing on the state of scientific publishing by saying, “Science advances through trust,” and he set the session’s focus on how fraud, commercial incentives and new technologies are changing the scholarly record.
The chair and witnesses described a fast-growing market for fabricated or low-quality articles produced by so-called paper mills and predatory journals. Kate Travis, managing editor of Retraction Watch, said the project’s public database lists more than 64,000 retractions and that retraction rates have risen substantially in recent years, with more than 5,000 retractions in 2025 alone. Travis said retractions are part of cleaning the record but stressed that detection and investigatory capacity are uneven across institutions and federal oversight offices.
Carl Maxwell, senior vice president for public policy at the Association of American Publishers, defended publishers’ role in curating and policing the record, saying publishers “invest their own money in the publication of each article” and warning that journal business models and immediate open‑access mandates create tradeoffs that must be managed.
Dr. Jason Owen Smith of the University of Michigan framed another angle of the problem: publishers are increasingly data firms that collect and sell analytics about how research and researchers are used. He warned that acquisitions such as a publisher purchase of an academic-hiring tool expand the firms’ access to sensitive administrative information and urged policymakers to treat data about federally funded research as a strategic public asset.
Across members’ questioning, lawmakers of both parties pressed witnesses on three near-term needs: (1) stronger and better‑resourced investigatory offices (ORI, NSF OIG) to detect and act on misconduct; (2) better metadata, open data and funding for reproducibility studies so errors are found and corrected before they propagate; and (3) careful policy design around open access and publication charges so the drive for immediate public access does not unintentionally erode peer review and quality checks.
The subcommittee left the record open for additional materials and asked witnesses to supply written follow-ups on specific allegations, the scale of publisher letters to journals, and technical approaches to detecting AI‑fabricated content.
