Google DeepMind Engineer Quits Over Rapid AI Development and Moral Concerns

Robert O'Callahan, who worked on a new generation of AI-accelerating chips at Google DeepMind, posted on X that he had submitted his resignation on Friday. He argued that making AI faster and cheaper is inherently irresponsible because the technology is progressing at a pace that outstrips safety safeguards. In a blog letter, he described his decision as a moral imperative rooted in his Christian beliefs, noting that ignoring the potential harms would not align with his faith.

O'Callahan, a New Zealand resident, said he will pursue AI-related work such as code-debugging agents but not in a way that accelerates risky capabilities. His exit follows similar departures by Anthropic’s Jacob Coxon and DeepMind researcher Bilal Chughtai, who also left over AI safety concerns. He clarified that his timing was unrelated to recent high-profile resignations, instead aligning with a ten-day backpacking trip on the Abel Tasman and Wangapeka tracks.