Fired OpenAI researchers say they were let go for 'prioritising safety'

Fired OpenAI researchers say they were let go for 'prioritising safety' Three former OpenAI researchers say their dismissals stemmed from raising safety concerns within the company. Mikita Balesni, Tomek Korbak and Jasmine Wang allege they were terminated after questioning internal practices they believed jeopardised the responsible development of AI systems. On 2 October a BBC‑reported spokesperson for the firm stated that an internal investigation “confirmed that these individuals mishandled delicate accusation extracurricular established institution procedures.” OpenAI later repeated the assertion in a message from its probe leaders, declaring, “We privation to beryllium precise wide that these decisions were not astir raising information concerns oregon speaking out,” and adding, “Our interior probe uncovered a important breach of spot beyond what's outlined in the missive they published and we basal by the determination to not proceed their employment.” The researchers contend the firings have created a chilling effect among staff. Balesni posted on X, “I judge we were fired for prioritising information implicit the near-term interests of OpenAI as a corporation,” alongside an open letter to the ChatGPT‑maker co‑authored with Korbak and Wang. In the letter the trio described themselves as former safety or alignment employees whose work sought to embed human ethical principles into AI technology. Korbak later reposted the letter, writing, “For months, I'd been raising information concerns that we're losing the quality to show what AI agents think, 1 of our champion tools for catching erstwhile they misbehave.” He followed with, “I judge that was wherefore I was fired.” Wang echoed the sentiment on X, saying, “We were not the archetypal to be pushed retired of OpenAI under suspicious circumstances,” and warned, “Unless the employees instrumentality a basal present against this benignant of manoeuvre, I am acrophobic we volition not be the last.” The dismissals occur amid a broader debate over AI safety and the potential societal risks posed by increasingly capable systems. OpenAI said it is finalising contracts with third‑party safety assessors and plans to disclose further details in the coming weeks. Read the full story at BBC News →