A Legislative Push for AI Guardrails
Senator Bernie Sanders (I-Vt.) and U.S. Representative Greg Casar (D-Texas) have introduced legislation aimed at curbing the advancement of artificial intelligence, specifically targeting systems defined as “artificial superintelligence.” The proposal, unveiled on Wednesday, seeks to permanently prohibit AI that exceeds human cognitive abilities or possesses the capacity to act independently of human control, according to the Austin American-Statesman.
The bill proposes the creation of a new Department of Artificial Intelligence to oversee the sector and would mandate a temporary pause on the development of advanced AI systems until federal safety guidelines are established. Sanders, speaking to the Associated Press, questioned the wisdom of developing technology that could eventually operate beyond human oversight. “Do we really want to develop a super intelligence that when it becomes smarter than human beings could act independently of human control?” he asked.
Defining the Stakes
The legislation explicitly targets “recursive self-improvement,” a process wherein AI models build more advanced iterations of themselves. As reported by HuffPost, the bill is supported by various industry researchers and engineers, including employees at Google DeepMind and OpenAI, who have expressed concerns regarding potential existential risks. These proponents argue that the current trajectory of AI development risks outrunning human capacity for safe management.
Beyond the existential risks, the legislation addresses the immediate economic impact of AI. Representative Casar, who chairs the Congressional Progressive Caucus, has been vocal about the role of AI in workforce displacement. Following layoffs at Oracle—which the company attributed to AI adoption—Casar has pushed for tighter scrutiny of how corporations utilize automation to replace human labor.
Political and Geopolitical Hurdles
Despite internal industry support, the bill faces significant opposition in the Republican-controlled Congress. President Donald Trump has resisted calls for increased oversight, recently stating at the United Nations that he intends to rebrand the technology as “super intelligence” and has shown little interest in curbing its growth. The administration’s stance is heavily influenced by the competitive landscape, particularly the race for technological dominance against China.
Sanders has argued that the risks posed by AI require a level of international cooperation similar to Cold War-era nuclear arms control treaties. He posited that avoiding catastrophic outcomes would necessitate a diplomatic approach between the U.S. and China to establish shared safety benchmarks, regardless of broader geopolitical tensions.