Wikipedia Officially Bans Use of AI in Writing and Editing Articles
In a decisive step highlighting shifts in the global digital landscape, Wikipedia (English version) has decided to almost completely ban the use of artificial intelligence in creating or rewriting its articles, after years of experiments and deliberations among the volunteer editor community.
The final decision came on March 20, 2024, with a landslide vote (40 in favor of the ban versus only 2 against), as editors considered that texts authored by AI often violate Wikipedia's fundamental policies, especially information accuracy and reliance on reliable sources; they may even add errors, biases, or unsupported content.
What does the ban include?
Complete ban: The use of AI to create or rewrite article content is prohibited.
Limited exception: Editors are allowed to use AI for language suggestions or editing texts written by the editor themselves, provided there is manual review and assurance that no new information is introduced "from" the AI or the meaning of the text is altered.
Translation: AI may be used for translation only if the editor themselves is fluent in both languages and can fully verify the accuracy of the translation.
Background of the decision: A long battle against synthetic texts
In recent years, Wikipedia faced a flood of "poor" AI-generated content, prompting its community to a series of corrective campaigns and failed experiments (including publishing automated summaries at the top of some articles, which were later removed under community pressure).
Many editors previously expressed cautious hope regarding the integration of AI, but they recently turned to real concern after administrative reports and incidents related to AI content increased to the point of becoming "exhausting," as one of them described.
Economic and political background as well
The ban comes after Wikipedia signed agreements with major technology companies (Amazon, Microsoft, Meta, Perplexity) due to these companies' use of the encyclopedia's content to train their large language models without compensation, leading to huge financial and technical burdens on the encyclopedia's servers.
The decision also comes as a clear step to differentiate from experiments such as the "Grokipedia" initiative, which produces all its content via the Grok chatbot and has been plagued by much criticism due to serious errors and unreliable sources.
Back to top button
Original source: TechWD
Comments (0)
Be the first to comment.