US and UK Lawmakers Take Coordinated Aim at 'Superintelligent' AI

In a significant move for global AI governance, legislators from the United States and the United Kingdom have introduced coordinated proposals targeting the most advanced forms of artificial intelligence. US Senator Bernie Sanders is preparing the "Ban Artificial Superintelligence Act," while UK MP Alex Sobel has tabled a similar motion in the House of Commons. Both initiatives share a core, ambitious goal: halting the development of AI systems that could surpass human-level intelligence.

The Core Demand: A Moratorium on Superhuman AI

The proposed bills outline a comprehensive regulatory approach that goes beyond a simple ban. Their key provisions include:

  • Development Ban: An outright prohibition on creating AI systems that reach or exceed the breadth of human cognitive abilities, often termed "superintelligent" AI.
  • Research Pause: A temporary halt on other advanced AI research until robust safety rules and governance frameworks are established.
  • Risk-Based Definition: A focus on systems that could "undermine state authority" or pose a threat to national sovereignty.

Global Ambitions: Treaties and Expanded Oversight

The lawmakers' vision extends beyond national borders. The proposals urge the US and UK governments to champion a global treaty banning the development of superintelligent AI worldwide. This represents a major push for a multilateral governance structure outside of existing UN forums.

Furthermore, the bills seek to grant governments new powers to monitor and restrict potential "precursor systems"—foundational models or research that could lead to superintelligence. This suggests regulators aim to intervene earlier in the development pipeline.

Drivers of the Movement: Warnings and Think-Tank Influence

This legislative push is backed by prominent voices from the AI community. Former OpenAI employee Daniel Kokotajlo and leading scholar Stuart Russell have publicly warned that unregulated AI advancement could lead to catastrophic infrastructure disruption or even a loss of human control. Their arguments provide a foundational rationale for the bills.

The think-tank ControlAI, which participated in drafting the legislation, described the proposals as a "starting point" for future comprehensive AI regulation, not a final solution.

Implications and the Road Ahead

Although both proposals are in their early stages, their symbolic and potential practical impact is substantial. This marks the first attempt within a G7 parliament to legally ban "superintelligent AI." It reflects deep-seated concerns among some policymakers about the exponential pace of AI progress and a proactive attempt to mitigate "singularity" risks.

The move is likely to fuel intense debate across the tech industry, policy circles, and academia, centering on the boundaries of innovation, national security, and global technological competition. Regardless of the outcome, the era of relatively unfettered AI exploration is facing increasingly concrete legislative challenges.