{
  "id": 10493736,
  "title": "When can we say AI made a scientific discovery?",
  "url": "https://urgent.news/2026/09/28/when-can-we-say-ai-made-a-scientific-discovery",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-28T17:03:16.000Z",
  "source": {
    "name": "MIT Technology Review",
    "slug": "mit-technology-review",
    "url": "https://www.technologyreview.com/2026/09/28/1145230/when-can-we-say-ai-made-a-scientific-discovery/"
  },
  "original_language": "en",
  "account": "Last Wednesday, Anthropic made headlines by introducing a molecular biology lab where Claude agents independently read and speculate about biological problems while human scientists perform experiments on their findings. This AI-powered lab claimed to have made its first discovery after 21 hours of work, which involved identifying a repeating pattern surrounding a known enzyme. The company emphasized that the pattern hadn't been catalogued before, and its resemblance to the gene-editing technology CRISPR suggested significance. However, some biologists argue that merely identifying a pattern, even if unique, is not a genuine scientific discovery. They contend that the crucial aspect is understanding what the pattern means – figuring out its function and potential applications. Furthermore, there are concerns that Anthropic's team might have been inspired by the work of Mario Rodríguez Mestre, a biologist at the University of Copenhagen, who had already discovered this pattern. Mestre expressed his suspicion to the New York Times, and Anthropic denied the accusation but stated that he has ceased using Claude. The controversy over AI's role in scientific discovery highlights the tension between presenting AI as an autonomous tool and recognizing the collaborative nature of scientific progress. While the feat of a general-purpose chatbot finding a distinctive pattern within a vast dataset is notable, the question remains whether this constitutes a genuine scientific breakthrough or merely an impressive demonstration of AI's capabilities.",
  "summary": "This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Last Wednesday, Anthropic announced that earlier this year it had launched a molecular biology lab, where Claude agents read and conjecture about hard biology problems and human scientists run experiments on what…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}