RFK Jr. Health Dept. Calls Nature ‘Junk Science’ and Cancels Subscriptions
In a controversial move that has reverberated across federal research institutions, the Department of Health and Human Services (HHS), under Secretary Robert F. Kennedy Jr., has officially canceled all subscriptions and contracts with Springer Nature, the publisher of the prestigious multidisciplinary journal Nature. Agency spokespeople confirm that biomedical researchers at NIH, USDA, NASA, and DOE no longer have access to Springer Nature titles—a decision Kennedy has framed as a stand against what he terms “junk science.”
Background and Rationale
On May 27, 2025, Kennedy stated during a podcast appearance that high-impact medical journals including The Lancet, New England Journal of Medicine (NEJM), and Journal of the American Medical Association (JAMA) are “corrupt,” accusing them of serving as “vessels for pharmaceutical propaganda.” He warned that unless these publications reform their peer-review and conflict-of-interest policies, NIH scientists would be barred from publishing and instead deposit findings in in-house, government-run outlets.
“Precious taxpayer dollars should not be used on unused subscriptions to junk science,” said Andrew Nixon, HHS spokesperson.
Budgetary and Technical Implications
Federal agencies collectively spend an estimated $5–10 million per year on enterprise licenses and individual journal subscriptions. According to USAspending.gov, DOE’s cancellation of its Springer Nature contract frees up approximately $1.2 million annually. NASA, which integrates Nature papers into mission-critical research pipelines—leveraging DOI-based APIs for automated literature mining in Terraformer and Earthdata workflows—will need to migrate to alternative sources or develop proprietary retrieval systems.
- Annual subscription fees per journal can range from $10,000 to $30,000, with bundled packages approaching seven-figure totals.
- Crossref and DataCite APIs currently power ~60% of DOI lookups in NASA’s Astrophysics Data System (ADS).
- Up to 80% of NIH grant proposals cite at least one Springer Nature publication in the preceding five years.
Open Access and Alternative Platforms
This shakeup accelerates the federal pivot toward open-access (OA) repositories. Agencies are now evaluating preprint servers such as bioRxiv and medRxiv, as well as institutional data repositories backed by NIH’s PubMed Central and DOE’s ScienceCinema. Under Plan S mandates in Europe, all research funded by coalition agencies must be OA by 2026. In response, Springer Nature has announced a new hybrid OA model—Nature Select—which offers immediate open access for a per-article processing charge of $9,500.
AI-Driven Literature Mining: New Strategies
With traditional subscriptions cut, agencies are exploring AI and machine-learning tools for semantic search and summarization over public corpora. Projects under the NIH’s Bridge2AI initiative and NASA’s Machine Intelligence for Science (MI-Sci) program leverage transformer models—such as BERT and RoBERTa—fine-tuned on PubMed abstracts to extract key findings, generate automated literature reviews, and detect dataset linkages.
- Custom NLP pipelines using spaCy and Hugging Face Transformers to parse PDFs and XML of OA articles.
- Knowledge graphs in Neo4j for mapping author–institution networks and funding relationships.
- Real-time alerts via Elasticsearch clusters hosted on AWS GovCloud, replacing proprietary alert services.
Expert Perspectives and Industry Response
Leaders in the publishing and research communities have voiced concern over the move:
- Dr. Alison Grey, Director of Open Science at the Chan Zuckerberg Initiative, warns that “stripping access to peer-reviewed literature undermines reproducibility and slows translational research.”
- Mark Oppenheimer, VP of Licensing at Springer Nature, notes that “government cancellations threaten the viability of rigorous peer review and the infrastructure supporting DOI resolution and metadata enrichment.”
Implications for Collaboration and Data Sharing
Loss of centrally curated publications could fragment collaboration—particularly in multi-agency efforts like the COVID-19 Rapid Response Consortium, where cross-referencing peer-reviewed findings was vital. Data-sharing frameworks such as FAIR principles (Findable, Accessible, Interoperable, Reusable) rely on persistent identifiers assigned by bodies like Crossref, an entity partly funded by membership fees from publishers including Springer Nature.
Conclusion
RFK Jr.’s subscription cancellations mark a significant inflection point in federal research policy. While proponents argue for budgetary prudence and a transition to open-access science, critics caution that eliminating access to established peer-review platforms jeopardizes scientific rigor, reproducibility, and international collaboration. As agencies develop in-house journals and AI-powered literature tools, the broader research ecosystem will be closely watching how quickly these alternatives can match the editorial quality, indexing standards, and infrastructure that Nature and its sister journals currently provide.