Gustavo Monnerat, Deputy Editor at The Lancet, shared on LinkedIn:
“Who checks the work when AI produces research faster than humans can verify it?
According to the 4th version of the European Commission’s Living Guidelines on generative AI in research (October 2026): you do.
My 8 takeaways:
- Humans stay accountable. Verify AI content, analyses and references.
- Disclose substantive use: which tools, what they did, and how you checked the output.
- Protect confidential material. Unpublished manuscripts, proposals and sensitive data stay out of tools that may reuse them.
- Think about reproducibility. Decide whether to keep prompts, logs and model versions.
- Check for errors and bias: fabricated references, wrong summaries of real papers, and answers that simply agree with you
- Institutions: provide training and secure tools
- Funders: set rules for AI in evaluation
- Use AI proportionately. Ask first whether it’s the right tool, including its environmental cost
Peer review, they stress, remains a fundamentally human process. Which point would you add?”
Other articles featuring Gustavo Monnerat on OncoDaily.