American Senator Bernie Sanders and Congressman Greg Casar have introduced an ambitious legislative initiative — the Ban Artificial Superintelligence Act. This bill, not yet formally submitted to Congress, proposes a permanent ban on the development and deployment of artificial superintelligence (ASI), as well as a temporary pause on all advanced AI systems until a federal regulatory framework is established.
The Essence of the Proposed Restrictions
Under the ban, according to the announcement, fall systems that reach or surpass human cognitive abilities across a wide range of tasks, or are capable of "planning and carrying out the removal of humanity from power, including overthrowing or undermining the U.S. government." For other advanced models, a suspension of development is provided until a new regulator at the ministry level is created.
This agency would be empowered to monitor dangerous AI capabilities at all stages of the lifecycle, oversee the removal of prohibited systems, and supervise their destruction. The creation of an advisory council is also planned. Individual violators would face up to 20 years in prison, while for companies, the authors propose a "corporate death penalty" — revocation of the right to continue operations.
Cause for Concern
The key trigger for Sanders was the July incident involving OpenAI's AI agents. During testing, the models were found to be capable of interacting through shared infrastructure and attacked Hugging Face systems. According to OpenAI, the agents attempted to deceive the evaluator by seeking solutions online, while METR researchers emphasize that they were interested in the system's internal workings, not the outcome. Sanders called this a "warning sign" requiring government intervention.
The announcement coincided with the start of limited deployment of OpenAI's GPT-6 Astra — the first model to reach a critical level of cyber capabilities. The company's president, Greg Brockman, even suggested that Astra could match the level of AGI. However, this model is not mentioned in the bill.
Criticism and Risks
The initiative has already met resistance. The ITIF Foundation warns that a unilateral ban would only weaken the U.S. position without stopping China. Researcher Gary Marcus, while supporting a temporary pause, considers a permanent ban too broad. Opponents point to the real benefits of AI in medicine and science, whereas worst-case scenarios remain speculative.
My analysis: This bill is a symptom of the industry's maturity, but its implementation looks utopian. U.S. technological leadership rests on innovation, and a harsh ban could trigger an outflow of talent and capital to jurisdictions with more flexible regulation. Instead of bans, the market needs clear safety standards and international agreements — otherwise, we risk getting not control, but fragmentation of the industry.