---
title: "When can we say AI made a scientific discovery?"
description: "An analysis argues that AI firms' claims of autonomous scientific breakthroughs obscure genuine progress and raise questions about how credit is divided between AI tools and human scientists."
kind: "wire"
section: "Machines"
author: "Josh Alley"
author_url: "https://hastalavista.ai/authors/alley"
published: "2026-10-07T00:53:39.077Z"
updated: "2026-10-07T00:53:39.077Z"
canonical: "https://hastalavista.ai/machines/when-can-we-say-ai-made-a-scientific-discovery"
source_url: "https://www.technologyreview.com/2026/09/28/1145230/when-can-we-say-ai-made-a-scientific-discovery/"
source_name: "MIT Technology Review"
archive_url: "https://web.archive.org/web/20261006010639/https://www.technologyreview.com/2026/09/28/1145230/when-can-we-say-ai-made-a-scientific-discovery/"
---

# When can we say AI made a scientific discovery?

*An analysis argues that AI firms' claims of autonomous scientific breakthroughs obscure genuine progress and raise questions about how credit is divided between AI tools and human scientists.*

By [Josh Alley](https://hastalavista.ai/authors/alley) · October 6, 2026

Source: [When can we say AI made a scientific discovery?](https://www.technologyreview.com/2026/09/28/1145230/when-can-we-say-ai-made-a-scientific-discovery/) — MIT Technology Review