MIT Technology Review reports OpenAI’s plan for an automated AI researcher

MIT Technology Review reports OpenAI’s plan for an automated AI researcher

On March 20, 2026, MIT Technology Review’s The Download reported that OpenAI had made a fully automated AI researcher its “north star” for the next few years. The newsletter describes this as an agent-based system intended to tackle large, complex problems by itself. OpenAI plans to build an “autonomous AI research intern” by September, for a small number of specific research problems; the newsletter says this would be a precursor to a multi-agent system slated to debut in 2028. In an interview, OpenAI chief scientist Jakub Pachocki discussed the plans with Will Douglas Heaven, who wrote the item. The distinction between a limited research intern and the more ambitious system matters: the stated plan is staged, and the larger system is a future goal, not a reported achievement. The newsletter does not describe how the systems would be evaluated, what research problems they would handle, or how much human oversight they would need.

Publication: 7397b79e-e4ce-b70f-941f-0b68c7eb682a

attribution — pass: The newsletter issue comments on several items in its own words, so it is one piece. The reading attributes the OpenAI item to MIT Technology Review's The Download and to Will Douglas Heaven, who signed it. It identifies Jakub Pachocki as OpenAI's chief scientist. OpenAI's goals are kept as reported plans.

In an exclusive interview this week, OpenAI’s chief scientist, Jakub Pachocki, talked me through the plans.

summary — pass: The title, summary and question fit the OpenAI item. The summary explains the 'fully automated researcher' as a system meant to tackle large, complex problems by itself. It sets out the staged plan: an intern by September, then a multi-agent system in 2028. It names the actor and the interview. The question follows from gaps the source leaves open. 'Multi-agent system' is not defined, but the meaning stays readable from context. The reading covers only the lead item of a multi-story newsletter, and its title makes that scope clear.

building an AI researcher—a fully automated agent-based system capable of tackling large, complex problems by itself

date — pass: The reading's date, March 20, 2026, matches the UTC day of originalPublished (2026-03-20T13:15:45Z). The September intern and the 2028 debut are presented as plans and slated goals, not outcomes, and there is no hindsight framing. The limitations disclose that this is an archived copy. The archive capture time is the same day, and the reading relies on no later sidebar material.

By September, the company plans to build “an autonomous AI research intern”

support — pass: Each claim matches the source text verbatim in substance. The source supports the point that the larger system is a future goal rather than an achievement, because it describes it as slated to debut. The noted gaps in evaluation, problem scope and oversight are accurate absences from the supplied text. No benefits or costs are invented.

The intern will be the precursor to the fully automated multi-agent system, which is slated to debut in 2028.

media — pass: No media was selected, and no verified media candidates were supplied, so there is no mismatch in media identity or relevance.