Sakana AI's latest AI Scientist-v2 model has launched a revolution in the scientific community. This advanced AI system not only enables the creation of scientific papers independently, but also successfully passed the double-blind peer review of the ICLR2025 seminar, which is shocking. From hypothesis to experimental design, to data analysis and paper writing, AI Scientist-v2 demonstrates its comprehensive scientific research capabilities without human intervention throughout the process.
The creative process of AI Scientist-v2 is completely autonomous, and Sakana AI confirms that no humans participate in the modification during the entire process. This AI system can not only make scientific assumptions, but also design experimental plans, write and optimize experimental code, execute experiments, analyze data, and produce high-quality visual charts. What’s even more amazing is that it can complete every detail of the entire paper from title to references, even typesetting and formatting.

Although the paper generated by AI successfully passed peer review, Sakana AI finally decided to withdraw the paper voluntarily. This decision reflects their high emphasis on academic ethics. At present, the academic community has not yet formulated clear norms on the publication of AI-generated content. Sakana AI hopes to promote the academic community to reach a consensus on the boundaries and norms of AI participation in scientific research, rather than hastily allowing AI scientists to publish papers.
This incident is not only a major breakthrough for AI in the field of scientific research, but also triggers profound thoughts on academic integrity, author identity recognition, and the boundaries of AI's participation in academic research. How to maintain academic innovation while ensuring research integrity will become an urgent issue that the scientific research community must face directly. In the future, with the continuous development of AI technology, the academic community will face more similar challenges, and it will become particularly important to develop clear norms and standards.