{"schema":"postcutoff/event@1","as_of":"2026-11-22T13:32:00+01:00","url":"https://postcutoff.com/e/2026-10-21-wikipedia-ai-cleanup-pangram-share/","md":"https://postcutoff.com/e/2026-10-21-wikipedia-ai-cleanup-pangram-share/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"human editor","policy":"https://postcutoff.com/about/"},"license":null,"id":"2026-10-21-wikipedia-ai-cleanup-pangram-share","date":"2026-10-21","date_precision":"day","short_title":"Wikipedia's AI Cleanup volunteers","deck":"Pangram flags ~15–16% of new English articles as AI-written in early 2026, down from ~20% in late 2025","takeaway":null,"category":"culture","category_label":"Culture","importance":2,"confidence":"medium","status":{"key":"partly","labels":["Partly confirmed"]},"sources":[{"n":1,"title":"IEEE Spectrum: How Wikipedia's editors keep the online encyclopedia human-written (Oct 21, 2026)","url":"https://spectrum.ieee.org/wikipedia-wikiproject-ai-writing-cleanup","type":"press","group":"press","domain":"spectrum.ieee.org"},{"n":2,"title":"Hacker News discussion (483 points)","url":"https://news.ycombinator.com/item?id=50257326","type":"discussion","group":"reactions","domain":"news.ycombinator.com"}],"official":0,"filed":"2026-10-28","updated":"2026-10-28","orgs":["Wikipedia","Wikimedia Foundation","Pangram Labs"],"title":"Wikipedia's AI Cleanup volunteers: Pangram flags ~15–16% of new English articles as AI-written in early 2026, down from ~20% in late 2025","summary":"In an Oct 21, 2026 IEEE Spectrum interview, Ellia Lebleu, a founder of English Wikipedia's volunteer WikiProject AI Cleanup, said Pangram's detector flagged about 20% of new articles and drafts as AI-written in late 2025 and about 15–16% in early 2026, after English Wikipedia barred editors from generating articles with LLMs. The Wikimedia Foundation is building a \"Paste Check\" for chatbot text and an attribution framework for AI users.","key_facts":["Source: IEEE Spectrum (Eliza Strickland), 'How Wikipedia's Editors Keep the Online Encyclopedia Human-Written', Oct 21, 2026; reached 483 points on Hacker News on Oct 27","WikiProject AI Cleanup started in 2023; Lebleu noticed the pattern around December 2023 (articles with 'conclusion' and 'Why is this important?' sections)","Pangram estimates (as reported): ~20% of new articles and drafts AI-flagged in late 2025, ~15–16% in early 2026; Lebleu calls them a conservative lower bound (no false positives on pre-ChatGPT text) and says the drop may reflect the policy or newer models Pangram has not adapted to","English Wikipedia prohibits using LLMs to generate articles (adoption date not given in the article); most AI text comes from beginner editors","Typical problem has shifted from invented citations to real sources that do not support the claim, which Lebleu links to retrieval-augmented generation","Tools: a volunteer-built AI source verifier; the Wikimedia Foundation's 'Paste Check' to flag text pasted from chatbots; a planned attribution framework showing AI users an article's last edit date, completeness and flagged problems","Lebleu: Claude has sometimes refused, saying 'I can't write a Wikipedia article; it would go against Wikipedia's policies'"],"key_numbers":[],"tags":["wikipedia","ai-generated-content","detection","moderation","pangram"],"science":null,"body_md":"## What happened\n\nIEEE Spectrum interviewed Ellia Lebleu, a founding member of English Wikipedia's WikiProject AI Cleanup, about how volunteers spot\nand remove machine-written text. Detection has moved from stock phrases (\"stands as a testament\", \"rich cultural heritage\") to\nstructural tells (the \"rule of three\", negative parallelisms) and output volume, such as five multi-thousand-word articles in five\nminutes. Pangram's figures, as relayed by Lebleu, put the AI-flagged share of new articles at about a fifth in late 2025 and a\nsixth in early 2026.\n\n## Why it matters\n\nWikipedia is one of the largest sources of training and retrieval data for language models, so how much model-written text it\nabsorbs feeds back into future models. The numbers are one vendor's detector, relayed in an interview, not an audited count.","disputed":[],"related":[{"id":"2026-06-02-neurips-2026-position-track-ai-papers","url":"https://postcutoff.com/e/2026-06-02-neurips-2026-position-track-ai-papers/","date":"2026-06-02","date_precision":"day","short_title":"NeurIPS 2026: 28% of position-track submissions score 100% AI-written, and 178 are desk-rejected","deck":null,"takeaway":"It was the first major ML venue to desk-reject papers at scale based on an AI-text detector.","category":"policy-safety","category_label":"Policy & safety","importance":2,"confidence":"high","status":{"key":"confirmed","labels":["Confirmed"]},"sources":1,"official":1,"filed":"2026-09-29","updated":"2026-09-29","orgs":["NeurIPS","Pangram Labs"]}],"people":[],"posts":[],"videos":[],"models":[],"changes":[{"date":"2026-10-28","type":"filed","text":"Created (Hacker News)"}],"provenance":{"agents":[{"model":"Claude Opus 5.5","maker":"Anthropic","tool":"Claude Code"}],"filed":"2026-10-28","run":null,"sources_read":null,"updated":"2026-10-28","human_review":null,"version":{"date":"2026-10-28"}},"gaps":[{"model_id":"gpt-6-2-astra","name":"GPT-6.2 Astra","cutoff":"2026-05","days_after":143,"in_training_data":false},{"model_id":"claude-opus-5-5","name":"Claude Opus 5.5","cutoff":"2026-06","days_after":113,"in_training_data":false},{"model_id":"gemini-4-hydrogen","name":"Gemini 4 Hydrogen","cutoff":"2026-03","days_after":204,"in_training_data":false},{"model_id":"grok-4-7","name":"Grok 4.7","cutoff":"2026-05","days_after":143,"in_training_data":false}],"short_url":null}