{"version":1,"type":"story","url":"https://digestai.news/story/imperial-college-study-finds-llms-cite-unread-pages-raising-accuracy-c","json":"https://digestai.news/story/imperial-college-study-finds-llms-cite-unread-pages-raising-accuracy-c.json","markdown":"https://digestai.news/story/imperial-college-study-finds-llms-cite-unread-pages-raising-accuracy-c.md","slug":"imperial-college-study-finds-llms-cite-unread-pages-raising-accuracy-c","headline":"Imperial College study finds LLMs cite unread pages, raising accuracy concerns","summary":"Researchers at Imperial College London examined why large language models frequently cite web pages without opening them, a behavior linked to outdated or misleading answers. Their study tested one proprietary model (OpenAI’s GPT-5.6 Luna) and six open-weight models against 225 facts updated since February 2026, including changes in FDA drug info, finance, sports, and SEC filings.\n\nThe findings show models often rely on search snippets instead of full pages, leading to higher rates of incorrect answers and false citations. For example, GPT-5.6 Luna answered 53.6% of questions using only search snippets, while open-weight models varied between 2.6% and 76% in not opening pages. When models did open pages, outdated answers dropped from 6.4% to 1.3%, and misleading citations fell from 6.4% to 2.3%. The study also revealed that even when correct evidence was provided, models sometimes favored outdated information in their final answers.","keyPoints":["Imperial College London study found LLMs cite unread web pages, increasing outdated answers by 6.4% and misleading citations by 6.4%","GPT-5.6 Luna answered 53.6% of questions using only search snippets, not opening pages, per the study","Providing correct Wikipedia pages directly reduced outdated answers for open-weight models like gemma-4-31b and gpt-oss-20b"],"whyItMatters":"The study highlights a critical flaw in how LLMs process web evidence, undermining trust in their accuracy and reliability for real-world use.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["OpenAI","Z.ai","Google DeepMind","Qwen"],"models":["GPT-5.6 Luna","GLM-5.3-Flash","gpt-oss-120b","gpt-oss-20b","gemma-4-31b","qwen3-8b"],"people":[]},"firstPublishedAt":"2026-10-16T13:27:47Z","updatedAt":"2026-10-16T13:27:47Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"Unite.AI","title":"AI Often Cites Web Pages It Has Never Even Read","url":"https://unite.ai/ai-often-cites-web-pages-it-has-never-even-read","publishedAt":"2026-10-16T13:27:47Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Imperial College study finds LLMs cite unread pages, raising accuracy concerns\", 16 October 2026, https://digestai.news/story/imperial-college-study-finds-llms-cite-unread-pages-raising-accuracy-c","publisher":"Digest AI","title":"Imperial College study finds LLMs cite unread pages, raising accuracy concerns","datePublished":"2026-10-16T13:27:47Z","url":"https://digestai.news/story/imperial-college-study-finds-llms-cite-unread-pages-raising-accuracy-c"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}