OpenAI on Friday doubled downward on its decision to fire three AI safety researchers, saying they were axed for mishandling delicate information, not for voicing concerns concerning the technology's risks.
In a new statement, the Sam Altman-led business stated it parted ways alongside Jasmine Wang, Mikita Balesni, and Tomek Korbak following an inner inquiry established they had "violated apparent policies on handling delicate information."
"Our inner inquiry uncovered a important breach of rely beyond what's outlined in the letter they published and we remain by the decision to not continue their employment," OpenAI said.
The AI lab added, "We desire to be extremely apparent that these decisions were not concerning raising safety concerns or speaking out." The business additionally stated that it encourages those types of conversations and considers them "essential to making the correct decisions."
The three fired researchers released a letter on social media a day earlier, addressed to OpenAI's safety committees, alert that their abrupt terminations would have a "chilling" consequence on business culture.
The letter, titled "OpenAI cannot create AI harmless on its own," stated that the ex-researchers have "become concerned that inner and external communications about our firing have made our erstwhile colleagues fearful to conversation and run in ways that, until final week, were an integral part of operating at OpenAI."
The trio wrote in the letter that AI is "not a normal technology" and that OpenAI is "not a normal company."
"Those of us who activity on safety see risks before anyone else, and we depend on near collaboration alongside exterior experts to activity out how to location them," the letter read. "The liberty to do so without fear, and to have well-defined inner procedures that allow this work, is itself an essential safety mechanism."
In its Friday response, the ChatGPT-maker stated it accepted that preserving the "monitorability" of frontier models requires industry-wide commitment, including from OpenAI.
"Monitorability has lengthy been a center part of our investigation program, and item we continue to allocate important resources in," the business said.
OpenAI stated it was "deeply sad" concerning the outcome involving the erstwhile researchers and praised their contributions to AI safety at the company.
"These decisions were not concerning them raising safety concerns," the business said. "We have continually encouraged that and continually will."
Read next
Natalie is a elder newsman on Business Insider's Business News squad covering the AI and data center boom.She was earlier on BI's Legal Affairs team, anywhere she handled important cases coming out of province and national court, as fine as bankruptcy. Her safety frequently focused on stories at the intersection of law, business, politics, and technology. Natalie has covered Donald Trump’s criminal and civil cases, the motion of lawsuits against the second Trump administration, the indictment and offender test of Sean “Diddy” Combs, the shooting death of UnitedHealthcare CEO Brian Thompson, and the lawful battles facing Elon Musk and his companies. Natalie came to Business Insider in June 2021 as a breaking news reporter, focusing on the most engaging angles on the day's trending news. Natalie mostly drove BI’s safety about the fatal “Rust” shooting involving Alec Baldwin and the disappearance and murder of Gabby Petito.Prior to joining BI, Natalie worked for the New York Post, the New York Daily News, and The Brooklyn Paper. She has an extended backdrop covering infringement and courts. During her additional than 12-year reporting career, she did a stint covering the constabulary attack out of the New York Police Department's headquarters. Natalie, a Brooklyn native, graduated from Brooklyn College in 2012 alongside a flat in journalism.
