Wikipedia will not allow artificial intelligence systems to directly edit articles on its platform because of concerns over accuracy and reliability, cofounder Jimmy Wales has said.
Wales explained that while AI technology has improved, the issue of fabricated responses, commonly known as AI hallucinations, remains a major challenge.
He said the risk of false information being confidently presented means AI cannot yet be trusted to make direct changes to Wikipedia content.
AI may help editors find information
Although Wikipedia is not ready to hand editing responsibilities to AI, Wales said artificial intelligence tools could still support the platform’s global community of editors.
He explained that AI agents could help identify important developments, especially smaller news stories or topics that may otherwise be overlooked by the millions of volunteers who contribute to Wikipedia.
AI companies rely on Wikipedia content
Artificial intelligence platforms frequently use information from Wikipedia when generating answers for users.
Wales noted that this has contributed to increased visits from AI bots, while traffic from human users has declined.
He described the drop in human visitors as significant but said it does not represent a major threat to the online encyclopedia because Wikipedia does not depend directly on advertising traffic.
Wikipedia calls for fair support from AI firms
The Wikimedia Foundation trustee said Wikipedia’s donation-based model means the platform relies on public support to operate.
Wales urged AI companies benefiting from Wikipedia’s information to contribute fairly, noting that handling millions of automated requests increases operating costs.
He added that Wikipedia has already reached agreements with some technology companies while taking steps against platforms that do not follow acceptable practices.