Wikipedia establishes detection guides and speedy deletion rule for AI-generated content

entered the ledger 19 Oct 2026 · last checked 19 Oct 2026 · Side effect

Wikipedia volunteer editors face a steady influx of undisclosed AI-generated articles and drafts characterized by Markdown artifacts, canned source hedging, and generic puffery. In response to the burden on volunteer maintenance, the community established a formal speedy deletion policy criterion (G15) for unambiguously LLM-generated pages and compiled a specialized field guide to help editors detect subtle AI prose.

The institution Wikipedia institution

The mechanism Friction removed externality scale 4 of 5, fast

Adaptation reform the rule itself was rewritten

The case as the ledger holds it

The assumption that broke: Writing encyclopedia articles requires manual human effort to locate and synthesize sources, which historically acted as a natural filter against low-effort mass generation.
The first-order effect: Unchecked influx of undisclosed LLM-generated text into Wikipedia drafts, articles, and edit summaries, requiring volunteer editors to manually spot AI tells and manage deletion processes.
Who pays: Volunteer editors and reviewers who spend time detecting, verifying, and cleaning up AI-generated text and hallucinated citations · group: Volunteers and reviewers
Scale and speed: 4 of 5 · fast

Evidence 1 page read

Wikipedia ↗
This is a list of writing and formatting conventions typical of AI chatbots such as ChatGPT, with real examples taken from Wikipedia articles, drafts, comments, and other content. It is a field guide to help detect undisclosed AI-generated content on Wikipedia

The second reader every claim, checked against the page

Claim	Verdict	Note
The speedy deletion policy criterion G15 (Unambiguously LLM-generated pages) lists some signs of AI writing, but is limited to the most objective ones.	supported	Page states: "The speedy deletion policy criterion G15 (Unambiguously LLM-generated pages) lists some signs of AI writing, but is limited to the most objective ones."
As of October 2026, the most common signs of newly AI-generated text on Wikipedia (besides the unambiguous indicators above) are Markdown presence, AI edit summaries, hedging disclaimers about source availability and usage, vague expression of connection or association, and canned emphasis on source attribution and media coverage.	supported	Page states: "As of October 2026, the most common signs of newly AI-generated text on Wikipedia (besides the unambiguous indicators above) are Markdown presence, AI edit summaries, hedging disclaimers about source availability and usage, vague expression of connection or association, and canned emphasis on source attribution and media coverage."
A high percentage of detected AI writing from the results of a detection tool is not a valid criterion for speedy deletion under G15.	supported	Page states: "A high percentage of detected AI writing from the results of a detection tool is not a valid criterion for speedy deletion under G15."
Another 2025 study on German theses has shown that humans managed a 'recognition rate of 57 % for AI texts and 64 % for human-generated texts'.	supported	Page states: "Another 2025 study on German theses has shown that humans managed a 'recognition rate of 57 % for AI texts and 64 % for human-generated texts'."
In a 2025 preprint, a subset of 'expert' annotators correctly identified 90% of AI-generated text (and misidentified 3.3% of human-written text) as AI-generated.	supported	Page states: "In a 2025 preprint, a subset of 'expert' annotators correctly identified 90% of AI-generated text (and misidentified 3.3% of human-written text) as AI-generated."