OpenAI Says Three Fired Safety Researchers Got The Boot Over ‘Breach Of Trust,’ Not Their Warnings

A frontier artificial intelligence firm says it fired safety researchers for breaking its trust, not for blowing the whistle.

OpenAI unveiled on Friday that an internal probe found “a significant breach of trust” by three former safety researchers: Mikita Balesni, Tomek Korbak and Jasmine Wang —hours after the group released a letter warning that the company’s handling of their firing has made colleagues afraid to speak up. This comes as the tech giant faces bipartisan pressure from the Senate with a potential bill to hold AI companies liable for agent hacking incidents and a copyright lawsuit from USA Today.

“Our internal investigation uncovered a significant breach of trust beyond what’s outlined in the letter they published and we stand by the decision to not continue their employment,” OpenAI said on X in a Friday note from its research leaders. “We have not and do not terminate any of our employees for raising concerns.”

The company said the research trio broke policies on handling sensitive information, according to the Friday note.

The ChatGPT maker also said it shares the researchers’ view that keeping frontier models must be monitored by the company and the industry, the note stated.

OpenAI referred the Daily Caller News Foundation to its Friday newsroom X post when asked for comment.

“I believe we were fired for prioritizing safety over the near-term interests of OpenAI as a corporation,” Balesni said in a Thursday post on X, sharing the researchers’ letter to company leadership.

“We have become concerned that internal and external communications around our firing have made our former colleagues afraid to speak and operate in ways that, until last week, were an integral part of working at OpenAI,” the letter states in part.

The researchers argued that superintelligence cannot be developed safely unless staff can collaborate openly within the company and with outside experts, according to the letter.