Academic journals' AI policies fail to curb the surge in AI-assisted academic writing

Current AI policies are ineffective at controlling adoption: Despite 70% of journals implementing AI usage guidelines, researchers' use of AI writing tools has increased dramatically across all disciplines, with no meaningful difference between journals that have policies and those that don't.

Transparency remains virtually non-existent: Of 75,000 papers published since 2023, only 76 (0.1%) explicitly disclosed their use of AI tools, revealing a massive gap between policy requirements and actual researcher compliance.

Vulnerable populations may be disproportionately affected: Non-English-speaking countries and researchers in physical sciences show the highest growth rates in AI tool usage, suggesting that language barriers and disciplinary differences may create unequal impacts across the global research community.

Current disclosure-based policies are insufficient: The study's analysis of over 5.2 million papers shows that simply requiring researchers to disclose AI use—the primary approach adopted by journals—has failed to promote transparency or meaningful accountability in academic publishing.

Urgent need for stronger ethical frameworks: The authors conclude that the current approach to governing AI in academia is inadequate and call for a fundamental re-evaluation of how the scientific community should integrate AI responsibly, moving beyond voluntary disclosure policies.