Issues have been expressed by MIT over a contentious report on how artificial intelligence affects innovation and research. The prestigious institute suggests that the paper should be removed from public debate over concerns linked to its "integrity".

The subject of the contentious study, "Artificial Intelligence, Scientific Discovery, and Product Innovation," was an MIT PhD candidate studying economics.. It purportedly illustrated how integrating an AI tool in a discrete yet significant materials science lab led to an increase in material discoveries and patent filings. However, it seemed to tamper with researchers' job satisfaction levels.

Esteemed MIT economists Daron Acemoglu, a recent recipient of the Nobel Prize, and David Autor both applauded the paper last year. In fact, Autor expressed to the Wall Street Journal that he was "floored". The duo, in an official MIT statement, recognized the paper as a topic of widespread discussion within the scholarly community on AI and science despite being unpublished in any peer-reviewed journals.

Yet, presently, both economists have lost faith in the origins, dependability, and the authenticity of the data, as well as the legitimacy of the research itself.

As reported by the WSJ, a computer scientist specialized in materials science shared his worries in January with Acemoglu and Autor. This resulted in an internal scrutiny by MIT.

MIT has chosen to keep the results of the probe under wraps due to student privacy rules. However, the paper's author is "no longer linked with MIT." The first press attention and a draft version of the paper both refer to the author as Aidan Toner-Rodgers, despite the institution's refusal to name him. Toner-Rodgers has not yet responded to TechCrunch's request for comment.

Also, the work was submitted for publication after MIT requested that it be withdrawn from The Quarterly Journal of Economics. The institute also requested for its removal from the preprint platform arXiv. Conventionally, only authors of the paper have the authority to submit withdrawal requests on arXiv, however, "as of now, the author hasn't initiated so", according to MIT.