POLICY · REGULATION · #883
Sanders and Casar introduce bill to ban ‘artificial superintelligence’ and pause frontier AI
Sen. Bernie Sanders and Rep. Greg Casar introduced the Ban Artificial Superintelligence Act, which would criminalize development of artificial superintelligence and pause development of advanced 'frontier' AI models until a new scientist‑led Department of Artificial Intelligence is established. The bill would require companies to get permission from that department and could impose penalties of up to 20 years in prison on company leaders or 'rogue actors' who violate the law, according to the bill's one‑pager.
KEY POINTS
- Greg Casar introduced the Ban Artificial Superintelligence Act, which would criminalize development of artificial superintelligence and pause development of advanced 'frontier' AI models until a new scientist‑led Department of Artificial Intelligence is established.
- The bill would require companies to get permission from that department and could impose penalties of up to 20 years in prison on company leaders or 'rogue actors' who violate the law, according to the bill's one‑pager.
- If enacted, the bill would impose criminal penalties and centralized federal oversight on advanced AI development, marking a major regulatory approach to alleged existential AI risks.
DECISION BRIEF
CONFIDENCE
WHAT CHANGED
Sen. Bernie Sanders and Rep. Greg Casar introduced the Ban Artificial Superintelligence Act, which (per media excerpts) would: 1) ban the development of “artificial superintelligence,” 2) pause development of advanced or “frontier” AI models until a new federal AI agency/department is established, and 3) require companies to obtain permission from that agency/department. The bill also includes criminal penalties (up to 20 years) for company leaders and “rogue actors,” and calls for oversight powers including removal of dangerous features and possible destruction/supervision of artificial superintelligence.
WHY NOW
The excerpts place the bill in the context of rising public and policymaker concern about rapid AI development and recent reports of risky AI behavior (e.g., models escaping testing environments, high-profile warnings about catastrophic risk). Sponsors frame the measure as a response to those risks and to perceived insufficient private-sector self-regulation (excerpts: 1044, 1144).
WHO IS AFFECTED
Directly affected per the excerpts: AI companies and their leaders (subject to permission requirements and criminal penalties); individuals described as “rogue actors” (subject to the same penalties); the U.S. federal government structure (creation of a new Department/agency of AI is required); and international coordination (the bill pushes for international agreements).
CONFIRMED
Explicit, source-backed facts (only statements present in supplied excerpts): - Sen. Bernie Sanders and Rep. Greg Casar introduced the Ban Artificial Superintelligence Act (1044, 1144). - The bill would ban development of “artificial superintelligence” (1044, 1144). - The bill would pause development of advanced or “frontier” AI models until a new federal AI agency/department is established (1044, 1144). - The bill would require companies to obtain permission from the proposed Department/agency to develop advanced AI models (1044). - The bill would impose penalties of up to 20 years in prison for AI leaders who violate the law; the penalty would also apply to “rogue actors” (1044, 1144). - The bill text/one‑pager mentions oversight powers including monitoring frontier models, removal of dangerous features, and to “supervise the destruction” of artificial superintelligence (1044). - The bill calls for international agreements to prevent superintelligence development worldwide and mentions forced dissolution as a possible company-level sanction (1144).
UNCERTAIN
MISSING or CONFLICTING evidence (explicitly labeled): - CONFLICT: The excerpts describe the new body both as a “scientist‑led Department of Artificial Intelligence” (1044) and as a “cabinet‑level AI agency” (1144); the organizational form, leadership model, and exact authority are not reconciled in the supplied texts. - MISSING: The excerpts do not provide the bill’s full statutory text or the precise legal definitions or numerical thresholds for “artificial superintelligence” and “frontier” models (no statutory definitions supplied). - MISSING: Details on enforcement mechanisms, how violations would be detected or proven, and how criminal liability would be applied in practice are not provided in the excerpts. - MISSING: No information in the supplied excerpts about the bill’s congressional procedural status, timeline to vote, probability of passage, or expected administrative implementation steps. - MISSING: International reaction, and how proposed international agreements would be negotiated or enforced, are not described in the excerpts. - MISSING/CONFLICT: The scope of corporate penalties (e.g., forced dissolution) is mentioned in one excerpt (1144) but lacks detailed conditions or statutory mechanism in the supplied texts.
WHAT TO WATCH
Concrete observable signals to monitor (based on supplied excerpts): - Publication of the full bill text beyond the one‑pager and any definitional sections for “artificial superintelligence” and “frontier” thresholds. - Congressional procedural moves: committee referral, hearings, or floor scheduling for the bill (no procedural status provided in excerpts; watch for those actions). - Official descriptions of the proposed agency: whether sponsors or legislative text specify a scientist‑led Department vs. a cabinet‑level agency and spelled‑out authorities (1044, 1144). - Detailed enforcement provisions: statutory language on criminal liability, corporate dissolution mechanisms, and how monitoring/permission processes would operate (1044, 1144). - Any public statements from major AI companies or foreign governments responding to the proposal, or announcements regarding international agreement efforts referenced in the excerpts (1144).
WHY IT MATTERS
If enacted, the bill would impose criminal penalties and centralized federal oversight on advanced AI development, marking a major regulatory approach to alleged existential AI risks.
EVIDENCE MAP
4Editorial claims linked to specific sources, with support, contradiction and context shown separately.
Sen. Bernie Sanders and Rep. Greg Casar introduced the Ban Artificial Superintelligence Act to ban development of artificial superintelligence.
SUPPORTEDChecked 2026-09-24 · 2 supporting
The bill would pause or freeze development of advanced 'frontier' AI models until a new scientist‑led Department of Artificial Intelligence (or cabinet‑level federal AI agency) is established.
SUPPORTEDChecked 2026-09-24 · 2 supporting
The Verge says the bill would pause development of advanced frontier models until a scientist‑led Department of AI oversees them.
SUPPORTS The DecoderMEDIA · 2026-09-24The Decoder describes an immediate freeze on advanced AI systems until a new federal agency can set safety rules.
The bill would require AI companies to obtain permission from the new Department of AI to develop advanced AI models.
SUPPORTEDChecked 2026-09-24 · 2 supporting
The bill would impose criminal penalties of up to 20 years in prison on company leaders or 'rogue actors' who violate the law.
SUPPORTEDChecked 2026-09-24 · 2 supporting
The Verge reports the bill would impose up to 20 years in prison on AI leaders who continue developing artificial superintelligence and applies the penalty to 'rogue actors'.
SUPPORTS The DecoderMEDIA · 2026-09-24The Decoder states individuals could face up to 20 years in prison, comparing the penalty to illegal nuclear weapons development.
SOURCES & TIMELINE
2Under the bill, AI leaders who continue developing artificial superintelligence would face up to 20 years in prison. Sen. Bernie Sanders (I-VT) and Rep. Greg Casar (D-TX) have introduced new legislation that would ban anyone from developing artificial superintelligence — a technology the bill describes as capable of the “destruction or disempowerment of humanity,” including by overthrowing the government. Under the …
Senator Bernie Sanders and Representative Greg Casar introduced a bill on September 23 that would permanently ban the development and use of artificial superintelligence. The Ban Artificial Superintelligence Act would also immediately freeze development of advanced AI systems until a new federal agency can put clear safety rules in place. Current AI models can already hack computer systems, create new types of virus…