CNRS warns AI-generated fraud is threatening scientific publishing integrity
IgorCarron · x · 2026-09-08
France's CNRS published a piece on generative AI's threat to research integrity: fabricated data, hidden prompts embedded in papers to manipulate AI reviewers, and absurd AI images (like a rat with a disproportionately large penis) are fueling the scientific publishing crisis. CNRS leans on open science and scientific integrity as counter-principles, with sleuths like Guillaume Cabanac tracking AI-generated fraud. The poster was unimpressed by CNRS's framing.
More from Safety
- When does a long-running agent become a security incident? Reddit debates the kill threshold — BlackMambla11 · 2026-09-08
- Internal AI bot leaked unannounced reorg plan and salary bands via over-scoped Drive access — Accomplished-Wall375 · 2026-09-08
- Researcher's Custom CTF Challenges Surprisingly Solved by GPT — terryyuezhuo · 2026-09-08
- Both Anthropic and OpenAI habitually read user transcripts for broadly defined 'safety' — aiamblichus · 2026-09-08
- Researcher argues AI pauses get easier with practice, not harder if used early — austinc3301 · 2026-09-08
- Developer pushes back on Codex privacy fears: user data almost certainly not used for training — _sholtodouglas · 2026-09-08