AI Hallucinations Plague Peer Review: Fake Papers Still Accepted by Top Conferences
gabriberton · x · 2026-08-03
The proliferation of AI-generated content is severely compromising the quality of academic peer review. A reviewer pointed out that during the ICLR 2026 review process, they flagged 11 papers for hallucinated references or authors, yet 2 were still accepted for oral spotlights.
Another researcher echoed this experience, noting they flagged a paper for hallucinations. The authors simply removed the hallucinated content in the revision, and the paper was ultimately accepted. This exposes vulnerabilities in top-tier conference review mechanisms when dealing with AI-generated fake content.
More from Safety
- Agent actions need checks outside the model: NVIDIA demos lean on guardrails amid wave of real agent incidents — Creamy-And-Crowded · 2026-10-16
- Vitalik builds "paranoid mode" agent wrapper with Tor, zkAPI and local prompt rewriting — DavideCrapis · 2026-10-16
- Security team demos Signal 1-click exploit technique at Hexacon Paris — moyix · 2026-10-16
- China moves to impose registration and licensing on frontier AI model weights and chip designs — deanwball · 2026-10-16
- Same agent task, two auth models: shared admin login wipes 1M audit rows, per-agent identity gets refused — Prompt Engineering · 2026-10-16
- Binding agent approvals to exact content: a synthetic test for stale permissions — Crescitaly · 2026-10-16