Sen. Bernie Sanders on Thursday called for a ban on the development of artificial superintelligence, arguing that the technology’s dangers were underscored by a recent interview with OpenAI’s chief executive. Sanders urged Congress to impose a moratorium on work that could produce systems exceeding human-level intelligence until robust safety measures, independent oversight and enforceable regulations are in place.
The statement intensifies a national debate over advanced AI after a high-profile interview in which OpenAI’s CEO discussed the capabilities and future trajectory of large-scale models. Sanders’ demand aligns with a chorus of voices from some lawmakers, researchers and advocacy groups pressing for stricter limits, while prompting warnings from parts of the tech sector that an outright ban could hinder beneficial innovation.
Sanders calls for ban on artificial superintelligence after OpenAI CEO interview, warning unchecked systems pose existential risk
Sen. Bernie Sanders on Wednesday urged Congress to move quickly to ban the development of artificial superintelligence, saying an interview with OpenAI’s CEO underscored how close industry leaders believe advanced systems may already be to capabilities that could outpace human control. In a forceful statement, Sanders warned that without robust limits and oversight these technologies could pose an existential risk to society, and outlined immediate policy steps he wants lawmakers to consider:
- an immediate moratorium on systems that exceed defined human-level control;
- mandatory independent audits of safety and alignment protocols;
- an international treaty to restrict development and deployment;
- significant federal funding for AI safety research and enforcement.
Political leaders, civil society groups and some technologists responded quickly, with advocates for stricter regulation applauding the move while industry voices cautioned against blanket bans that could stifle beneficial innovation. Legal analysts say the proposal sets up a test for whether Congress will pursue emergency legislation, hearings and cross-border coordination – a process that could reshape how the United States balances innovation, national security and public safety as AI capabilities accelerate.
Senator urges global moratorium and binding international treaty to halt development until independent safety audits and alignment benchmarks are met
In the wake of a high-profile interview with OpenAI’s CEO, Sen. Sanders pressed for immediate action, urging a global moratorium on artificial superintelligence development and a binding international treaty to govern progress. The senator called for work to halt until clear, third‑party safety verifications are completed and formal alignment benchmarks are met, arguing that voluntary corporate measures are insufficient to protect public safety and democratic institutions.
- Independent audits: certified third-party reviews of training data and model behavior
- Alignment benchmarks: measurable safety thresholds for deployment
- Transparency requirements: public reporting and access for regulators
- Enforcement: international sanctions and halt mechanisms for non-compliance
| Item | Minimum requirement | Indicative timeline |
|---|---|---|
| Independent audit | Third‑party certification | 6 months |
| Alignment benchmark | Safety ≥95% on tests | 12 months |
| Development pause | No scaling until criteria met | Until compliance |
Lawmakers and experts are now weighing how an international accord could be negotiated and enforced, with the senator advocating for rapid coordination through the United Nations and multilateral forums. Supporters say a treaty with clear inspection regimes and enforceable penalties would create global norms and prevent a regulatory race to the bottom; critics warn of stifling innovation, but the senator maintained that public safety must take precedence until verifiable safeguards are in place.
Sanders pushes Congress to hold hearings, mandate transparency, require emergency safety standards and fund alignment research to prevent runaway AI
Sen. Bernie Sanders said lawmakers must move quickly after a high-profile interview with OpenAI’s CEO exposed gaps in oversight and accountability for advanced machine-learning systems. He called for immediate action that would force companies and agencies to reveal safety testing, model capabilities and deployment plans, and sought a legal backstop to prevent unchecked development of systems that could behave unpredictably. His list of priorities included:
- Congressional hearings: public sessions to question executives, regulators and independent experts.
- Transparency mandates: disclosure of training data provenance, safety evaluations and risk assessments.
- Emergency safety standards: enforceable requirements for pause, rollback and response mechanisms.
- Alignment research funding: sustained grants to develop provable methods that keep models aligned with human interests.
Sanders outlined draft language for legislation that would pair short-term regulatory tools with multiyear research investments and even suggested a temporary moratorium on attempts to create systems that could surpass human cognitive control until safeguards are proven. Lawmakers of both parties reacted cautiously, citing the need for technical review; agency officials signaled readiness to hold expedited rulemaking if Congress acts. The proposal’s quick-reference implementation plan is summarized below:
| Measure | Target timeline | Indicative funding |
|---|---|---|
| Public hearings | 30-60 days | – |
| Emergency standards | 90 days | – |
| Alignment research program | 2-5 years | $1 billion |
In Summary
As Sanders pressed for a ban on the development of artificial superintelligence in the wake of the OpenAI CEO’s interview, he injected fresh urgency into an already contentious national conversation about how to govern powerful AI systems. His call – framed around risks to public safety, economic security and democratic institutions – is certain to deepen the divide between lawmakers seeking strict controls and industry leaders who argue that innovation, oversight and targeted regulation are better paths.
What happens next will depend on whether Sanders can marshal broader political support and translate his warning into concrete proposals such as hearings, legislative text or regulatory action. For now, the exchange has highlighted how rapidly AI policy debates are evolving and set the stage for a potentially pivotal clash between technologists, regulators and policymakers over the future limits of artificial intelligence.
