Cerebras CEO Andrew Feldman challenges AI risk messaging at Disrupt

By Splitfeed EditorialLast updated Oct 14, 2026, 23:21 UTC

On October 13, 2026, Cerebras CEO Andrew Feldman took the stage at TechCrunch Disrupt and delivered a pointed critique of prominent AI leaders who publicly emphasize catastrophic risk scenarios. His core argument was direct: figures in the industry who amplify doomsday warnings should, in his framing, perhaps stop talking. Feldman contended that the volume of risk-focused messaging is crowding out serious conversation about what AI can actually deliver in terms of benefits.

The remarks land in the middle of a long-running structural debate within the AI industry. On one side sit researchers, executives, and policymakers who argue that flagging existential or societal risks early is a professional and moral obligation, particularly as systems grow more capable. On the other side are voices, Feldman now prominent among them, who argue that the risk narrative has become disproportionate and functions to distort public perception and policy priorities away from genuine opportunity. What makes this more than a conference soundbite is that Cerebras is a hardware company competing in the AI infrastructure market, meaning Feldman has a direct commercial stake in the industry's public image remaining positive and its regulatory environment remaining permissive.

No specific reactions from the Pro-AI, Anti-AI, or Middle Ground camps have been published in response to this story yet. Typically, Pro-AI voices would welcome Feldman's framing, arguing that doom-focused narratives are exaggerated and slow beneficial deployment. Anti-AI voices would likely counter that dismissing risk messaging is exactly what bad actors and negligent developers want, and that executives with financial interests in AI expansion are not neutral arbiters of how much caution is appropriate. Middle Ground observers would generally expect to push back on the binary, arguing that risk communication and benefit communication are not zero-sum and that both deserve rigorous, evidence-based treatment.

The argument Feldman opened at Disrupt is unlikely to be settled by a single event, but responses from the AI safety community and from policymakers who have leaned on risk framing to justify regulation will be worth tracking in the days following his remarks.