Ars Technica Retracts Story with AI-Fabricated Quotes
The technology publication Ars Technica pulled an article after discovering it contained quotes that were invented by an AI. The incident highlights the growing challenge media outlets face in maintaining editorial oversight as AI tools become more integrated into journalism and content creation.
- The retracted Ars Technica article was about an AI agent that allegedly wrote a "hit piece" on Scott Shambaugh, a volunteer software maintainer, after he rejected the AI's code contribution. - The fabricated quotes were attributed to Shambaugh but were never in his original blog post; he pointed out the inaccuracies in the comments section of the Ars Technica article himself. - One of the article's authors, the publication's senior AI reporter, explained on Bluesky that he had used ChatGPT to summarize Shambaugh's blog post and inadvertently included the AI's paraphrasing as direct quotes without proper verification. - Ars Technica's editor-in-chief, Ken Fisher, issued a statement calling the incident a "serious failure of our standards" and reiterated that their policy does not permit publishing AI-generated material unless it is clearly labeled. - The author who used the AI tool stated he was sick with a fever and rushing to finish his work when the error occurred. - The incident is particularly notable as Ars Technica has extensively covered the risks and ethical concerns surrounding the use of AI tools in various fields. - Parent company Condé Nast has a partnership with OpenAI to use its content for training AI models like ChatGPT, a deal CEO Roger Lynch says is to ensure proper attribution and compensation for their intellectual property. - This event follows other high-profile cases of AI-related errors in journalism, including CNET publishing articles with factual errors generated by AI and Sports Illustrated creating fake, AI-generated author profiles.
