Economist Calls for Zero-Tolerance Ban on AI-Written Papers

Economist Paul Novosad posted a series of threads on 09-24 as "a disheartened reviewer," arguing that academic journals should adopt a zero-tolerance policy toward AI-generated writing, sparking debate in academia.

Confirmed

• Novosad's core claim is that "writing is thinking": journals are fully justified in requiring submitters to demonstrate that at least one scholar has genuinely thought deeply about the topic and written several pages in their own words.
• As a concrete measure, he suggests screening submissions with AI detection tools like Pangram, with violators banned from submitting for two years.
• He acknowledges some plausible legitimate uses of AI, but believes these exceptions will be drowned in a flood of low-quality AI-generated text, demoralizing reviewers and grinding journals down.
• He also notes that even if AI-generated economics research has value in itself, once it enters the existing journal system—under a mutual "don't ask, don't tell" understanding about AI involvement—peer review and the journal system will collapse; AI-generated research needs its own separate publishing venue.

Why it matters

• "Bad money drives out good" is the mechanism at the heart of the argument: unrestricted, diligent authors compete with low-quality AI-written submissions, the former gradually exit, reviewer motivation erodes, and journal quality declines systemically.
• The argument shifts the debate from "can we detect AI" to "do journals have the right to require authors to prove human authorship," touching on the very premises of peer review and proposing an alternative model for academic publishing in the AI era—separate venues for human-written papers and AI-generated research.

2026-09-24 ~ 2026-09-24 · 8 related posts

• [source] Economist Calls for Zero Tolerance on AI Writing in Journals as Slop Threatens Peer Review — paulnovosad · 2026-09-24
• Economist urges journals to adopt zero tolerance for AI-written papers, with 2-year bans — paulnovosad · 2026-09-24
• Referee laments AI-written papers: journals can require proof that authors actually think — paulnovosad · 2026-09-24
• [source] Economist: AI-assisted research will break the journal system's 'don't ask don't tell' equilibrium — paulnovosad · 2026-09-24
• Economist urges journals to adopt zero tolerance for AI writing — Franco_Calabres · 2026-09-24
• Economist on journal AI-writing bans: human ghostwriting was never the issue — paulnovosad · 2026-09-24
• Economist: AI submissions burn massive human review time to be identified as trash — paulnovosad · 2026-09-24

1 near-duplicate retellings: paulnovosad