OpenAI announced that, as part of its goal of using AI more independently in scientific research, it had developed an “automated research intern” capable of carrying out tasks whose boundaries and purpose are defined in advance under human guidance. The system can complete work that could take an experienced researcher several days. In this way, the company achieved on time its goal of a research assistant at intern level set for September 2026. OpenAI said it was also making progress toward developing a more independent AI researcher by March 2028, but that this stage had not yet been reached. CEO Sam Altman announced the goals during a livestream in October 2025. The company said that responsible use could directly benefit people. The system is not described as an independent researcher; rather, it refers to transferring certain tasks to AI instead of removing human researchers from the process entirely.
Why it matters
This step shows that AI is positioned in scientific work not as an actor making decisions on its own, but as an assistant taking on tasks whose scope is defined by human researchers. Since the issue is not the complete elimination of human labor, but the transfer to the system of specific time-consuming tasks handled by experienced researchers, the division of responsibility and forms of oversight within research teams are gaining importance. However, the level reached does not mean the same thing as the company’s goal of a more independent researcher; which tasks the system is capable of handling adequately and where human intervention remains necessary are still open questions. OpenAI’s timeline frames this progress as an interim stage in a gradual research automation plan; it also states that the more advanced stage has not yet been achieved.
Background
OpenAI is not a new name in the FikirPilot archive: we have published 22 news stories mentioning this name in the past 90 days; the latest was dated September 6, 2026.