Bernie Sanders Pushes Superintelligence Ban as Jacob Coxon Warns AI Labs Are 'Gambling With Our Lives'

Senator Bernie Sanders plans to introduce legislation to ban artificial superintelligence and pause advanced AI development, following Anthropic researcher Jacob Coxon's resignation and warning that leading AI labs are 'racing straight to self-improving superintelligence and gambling with our lives.' Coxon, who previously worked at OpenAI, said neither company is acting responsibly. Anthropic's Alignment Science Lead Evan Hubinger responded that he personally believes there is a >10% chance AI could kill all humans within the next decade.

Bernie Sanders Pushes Superintelligence Ban as Jacob Coxon Warns AI Labs Are 'Gambling With Our Lives' Coxon warned that leading AI labs were racing towards self-improving systems after leaving Anthropic, while Sanders used the warning to reinforce his call for a superintelligence ban Senator Bernie Sanders is pledging to introduce legislation to ban artificial superintelligence and pause advanced AI development, after Anthropic researcher Jacob Coxon resigned and accused leading labs of ' racing straight to self-improving superintelligence https://www.ibtimes.co.uk/ai-governance-expert-warns-recursive-self-improvement-1818039 and gambling with our lives'. 'Mr Coxon is right,' Sanders wrote on X on 9 September, quoting Coxon's resignation post. 'The very people building this technology admit that it could threaten the future of humanity https://www.ibtimes.co.uk/openai-chief-scientist-warns-ai-advancement-security-risks-1818185 .' 'That is why I will soon be introducing legislation to ban superintelligence and pause AI development.' Coxon said he had spent the previous three years doing pretraining research at OpenAI and Anthropic before resigning from Anthropic. 'Neither company is acting responsibly,' he wrote. 'They are racing straight to self-improving superintelligence and gambling with our lives.' Sanders and Representative Greg Casar had already announced the Ban Artificial Superintelligence Act on 3 September. Sanders' post gave fresh political weight to legislation unveiled six days earlier rather than introducing a new proposal prompted by Coxon's departure. Jacob Coxon Resigns From Anthropic Coxon's claims are allegations by a departing researcher. They do not establish that Anthropic or OpenAI broke the law, nor do they show that either company's AI systems pose an imminent threat. Coxon's resignation came after he had worked at both OpenAI and Anthropic on pretraining research. In his public post, he warned that developers were moving towards self-improving superintelligence despite what he described as serious concerns about the consequences. Another Anthropic researcher soon responded directly to Coxon's warning. Evan Hubinger Puts Personal Extinction Odds Above 10% Evan Hubinger, Anthropic's Alignment Science Lead, responded directly to Coxon's post. 'Jacob is correct here—we really do earnestly believe AI could kill all humans ' Hubinger wrote. 'I personally think it is 10% within the next decade.' Hubinger added that he believed Anthropic was 'trying its best', but said the company did not yet have a plan to solve alignment for superintelligence and was 'not clearly on track to' do so. The comments came from Hubinger personally and should not be treated as a formal statement by Anthropic's board or corporate leadership. His estimate was a personal assessment, not an established prediction. Alignment research focuses on making advanced AI systems behave in accordance with human goals and constraints. What Sanders' Bill Would Do The legislation Sanders and Casar announced on 3 September https://www.ibtimes.co.uk/ban-superintelligent-ai-legislation-1817896 would permanently ban the development and deployment of artificial superintelligence while imposing a temporary pause on advanced AI development until a new federal body establishes rules and safety requirements. The proposal would create a cabinet-level agency with powers to monitor advanced AI systems, oversee the removal of dangerous capabilities and supervise the destruction of artificial superintelligence. Its definition of superintelligence is broad. The lawmakers' summary describes systems capable of matching or exceeding human cognitive performance across a broad range of domains, as well as systems capable of planning or executing the disempowerment of humanity, including by undermining or overthrowing the US government. The proposal also includes severe penalties. Companies that violate its restrictions could face what Sanders and Casar describe as a 'corporate death penalty', while individuals could face up to 20 years in prison https://www.ibtimes.co.uk/us-bill-20-year-prison-superintelligent-ai-1817847 . The lawmakers compare the proposed penalties with those associated with illegally developing nuclear weapons. The bill also calls for international agreements to prevent the development of artificial superintelligence abroad, including measures involving export controls. Those are proposed provisions. The legislation has not become law, and none of its penalties are currently in force. Sanders' AI Ban Gets a New Political Argument Coxon publicly accused OpenAI and Anthropic of racing towards self-improving superintelligence. Hubinger, still working at Anthropic, separately said he believed AI could kill all humans and that Anthropic was not clearly on track to solve alignment for superintelligence. The underlying risk remains contested. Hubinger's estimate is his personal assessment, not an established prediction, and there is no scientific consensus that superintelligent AI will cause human extinction. Sanders is now using warnings from researchers working inside and around leading AI labs to make the case for a permanent superintelligence ban and a temporary pause on advanced AI development. © Copyright IBTimes 2026. All rights reserved.