Bernie Sanders, the veteran Vermont senator and two-time Democratic presidential candidate, wants to hit the brakes on artificial intelligence before it outpaces human control. His proposal calls for a legislative freeze on the development of advanced AI models until Congress enacts credible safeguards. It is, by any measure, the most aggressive legislative response to the AI boom in Washington since large language models became a fixture of daily life.
What Sanders Is Actually Proposing
Sanders is not calling for an outright ban on all AI. The bill he is co-introducing with Representative Greg Casar targets specifically what the proposal calls artificial superintelligence: language models that continuously improve without meaningful human oversight, to the point where they could surpass human-level intelligence and stop responding to shutdown commands. The formal name is the Artificial Superintelligence Ban Act.
The penalties in the draft legislation are steep. Anyone attempting to circumvent the restrictions could face up to 20 years in prison. Sanders drew an explicit parallel to the international frameworks built to limit nuclear weapons development, framing superintelligent AI as a comparable systemic threat.
Johana Bhuiyan
The senator names three companies he sees as the primary threat to political, social, and economic stability: OpenAI, Anthropic, and Meta. He has pointed to the growing chorus of AI researchers and developers who have themselves called for a slowdown, arguing that the technology is advancing without any meaningful pause for safety review. Sanders doesn't consider his position extreme. Only strict rules, in his view, can rein in what he calls the technological oligarchs of Silicon Valley.
The legislation would pause superintelligence development until a new federal agency, dedicated specifically to AI monitoring and governance, is fully operational and has established binding rules, review procedures, and a complete safety framework.
Beyond the Ban: International Coordination
Sanders and Casar are not trying to kill AI innovation outright. Their stated goal is to reshape how the technology develops, not to erase it. The pause would remain in effect only until the proposed federal oversight body is ready to enforce meaningful standards.
The bill goes considerably further than the EU's AI Act, the most comprehensive regulatory framework currently in force. Sanders explicitly calls for international cooperation: the United States, in his framing, should lead a coalition of allied nations to conclude binding agreements that limit uncontrolled superintelligence development globally, not just domestically.
The systems Sanders fears are not today's chatbots. He is focused on future models capable of toppling governments, evading shutdown commands, and operating beyond any human-defined boundary. Whether those risks are imminent or speculative is a question that genuinely divides the AI research community.
What It Means for DeFi and Crypto
The proposal has generated little enthusiasm in Washington. The United States Congress is currently consumed by more immediate crises: a volatile executive branch, a widening federal deficit, the uneven results of tariff policy, and ongoing conflicts in the Middle East. The Artificial Superintelligence Ban Act is unlikely to advance quickly in that environment.
If the bill were to gain traction, the DeFi sector would feel its effects. The crypto and decentralized finance community is genuinely split on where AI fits in their world. One camp sees unchecked AI development as a direct threat to blockchain security and exchange infrastructure, a spiral that could go dangerously off the rails. That group would likely welcome Sanders' proposal.
On the other side sits Vitalik Buterin, co-founder of Ethereum. As reported by SpazioCrypto, Buterin has argued that cryptocurrencies should not feel threatened by AI, pointing to the 15 years of cybersecurity engineering that has hardened Bitcoin and other major protocols. His position is that those same engineers can evolve system defenses to meet whatever threats AI eventually generates.
The real risk of a blanket ban cuts both ways. Suppressing AI development would constrain not only the genuine dangers of uncontrolled proliferation, but also the financial and technological opportunities the technology is opening up. For DeFi specifically, AI-driven smart contract auditing, fraud detection, and on-chain analytics represent a class of tools that would stall under a development freeze. The debate Sanders has forced into the open is one the crypto sector will need to engage with directly, whatever happens to the bill itself.
