Ban Artificial Superintelligence Act of 2026
- Sponsor
- Rep. Casar, Greg [D-TX-35]
- Committees
- Oversight and Government Reform Committee (primary)
- Last action
- Sep 24, 2026
Bottom line
The bill creates a new federal department to enforce a mandatory pause on advanced AI development and a permanent prohibition on artificial superintelligence, backed by stringent penalties and asset forfeiture.
What it actually does
This bill establishes a new Department of Artificial Intelligence (DAI) to regulate advanced AI systems. It mandates a pause on the development and deployment of advanced AI until the DAI is fully staffed and has promulgated safety rules. Crucially, it prohibits the development, deployment, acquisition, possession, funding, import, or transfer of artificial superintelligence (ASI) or systems exhibiting superintelligence precursor characteristics, with severe penalties including criminal charges, employment bans, and forfeiture of intellectual property and assets for violations.
Proponents argue
Proponents argue that the bill is a necessary and urgent response to the existential threats posed by unchecked artificial intelligence development, citing numerous warnings from leading AI experts and industry leaders about the potential for catastrophic consequences, including human extinction or loss of control. They contend that a proactive, robust regulatory framework, including a pause and outright prohibition on superintelligence, is essential to safeguard humanity.
Opponents contend
Opponents would likely argue that the bill's definitions of 'advanced artificial intelligence system' and 'artificial superintelligence' are overly broad and could stifle beneficial innovation and research. They might also contend that the severe penalties, including intellectual property forfeiture, are draconian and would drive AI development out of the United States, undermining national security and economic competitiveness. Concerns about governmental overreach and the practical challenges of enforcing such prohibitions on rapidly evolving technology would also be raised.
The bill is relatively short and clearly written, making it accessible for an informed reader to understand its core provisions and implications within a reasonable timeframe.
Section 4
Establishment of the Department of Artificial Intelligence (DAI)
This section establishes a new executive department within the U.S. government, named the 'Department of Artificial Intelligence.' Its stated purpose is to protect the public from AI dangers, specifically by enforcing a pause on advanced AI development and prohibiting artificial superintelligence.
Supporters argue
Supporters argue that a dedicated federal department is essential to centralize expertise and authority for regulating AI, given the technology's profound and potentially catastrophic implications. They believe existing agencies lack the specific mandate and resources to address AI's unique challenges effectively.
Critics contend
Critics might contend that creating an entirely new department is an overreaction and could lead to bureaucratic bloat and inefficiency. They may suggest that existing agencies could be tasked with AI oversight, or that a more agile, interagency task force approach would be more appropriate for a rapidly evolving field.
Tradeoffs
The tension lies between the perceived need for a robust, dedicated regulatory body and concerns about creating a large, potentially slow-moving bureaucracy that might struggle to keep pace with technological advancements.
Section 8
Mandatory Pause on Advanced Artificial Intelligence Development
This section imposes a mandatory pause on all advanced artificial intelligence systems, meaning they cannot be trained, modified, or fine-tuned (except for deactivation) until the new Department of Artificial Intelligence is fully staffed and has established clear rules for safe development and deployment. Additionally, no unreleased advanced AI systems may be deployed during this pause period.
Section 9(a)
Prohibition on Artificial Superintelligence (ASI) and Precursor Characteristics
This section outright prohibits any person from developing, deploying, acquiring, possessing, funding, importing, or transferring artificial superintelligence (ASI) or any AI system that exhibits 'superintelligence precursor characteristics.' This includes any components sufficient to reconstruct such systems' capabilities. ASI is defined as an AI system exceeding human cognitive performance across most domains or capable of planning humanity's destruction or disempowerment.
Section 10
Actions Against Systems Exhibiting Superintelligence Precursor Characteristics
This section mandates that the Secretary of Artificial Intelligence take immediate action against any AI system identified as exhibiting 'superintelligence precursor characteristics.' Such systems must be subjected to a mandatory pause and sequestered from the internet. If the Secretary cannot verify that the system no longer exhibits these characteristics within 30 days, it must be rendered inoperative. Any identified artificial superintelligence must be immediately rendered inoperative.
Section 12
Charters for Advanced Artificial Intelligence Development
This section establishes a chartering system, requiring any person developing or distributing an advanced artificial intelligence model to obtain a charter from the Secretary of Artificial Intelligence. As a condition of maintaining this charter, developers must agree to grant the Secretary full access to their AI systems, processes, staff, and physical infrastructure for enforcement purposes. Critically, if a charter is revoked due to a violation, the person forfeits their intellectual property and assets to ensure the destruction of related systems and hardware.
Section 13
Penalties for Violations
This section outlines severe penalties for violating the bill's provisions. Covered policymaking individuals or 'rogue actors' who recklessly violate sections 8, 9, or 12 face criminal fines and imprisonment for up to 20 years. Non-policymaking individuals face a 10-year ban from employment in the AI industry, with a waiver option for whistleblowers. Charters for covered entities will be revoked for violations, leading to forfeiture of intellectual property and assets to ensure the destruction of related systems, subject to appeal to the U.S. Court of Appeals for the Federal Circuit.
Section 15
International Coordination and Policy
This section directs the Secretary of Artificial Intelligence, in coordination with the Secretary of State, to pursue international agreements and allied coordination to prevent the global development of artificial superintelligence. It also establishes a U.S. policy to prevent ASI development worldwide, which may include the use of export controls on artificial intelligence computing infrastructure.
The computing power threshold for defining an 'advanced artificial intelligence system' (initially 10^25 operations) will be adjusted annually by the Secretary to reflect changes in training efficiency or other technological developments, maintaining equivalency.
Section 3(1)(B)
Why it matters:This is standard legislative practice for technical definitions that require future flexibility, ensuring the definition remains relevant as technology evolves without requiring new legislation.
Case for: Proponents would argue this flexibility is crucial for the law to remain effective in a rapidly changing technological landscape, preventing the definition from becoming quickly outdated.
Case against: Critics might argue that granting the Secretary unilateral authority to adjust such a critical threshold could lead to arbitrary changes or expand the scope of regulation without direct congressional oversight.
Estimated impact: This provision ensures the regulatory scope of 'advanced artificial intelligence systems' can dynamically adapt to technological progress.
One of the 'superintelligence precursor characteristics' is defined as 'The capacity to ensure continued and independent operation notwithstanding attempts to shut down or otherwise hinder operations.'
Section 3(6)(C)
Why it matters:This is a technical detail within a broader definition, which is common. However, its inclusion is highly significant for the bill's enforcement mechanisms.
Case for: Proponents would argue that an AI system's ability to resist shutdown is a critical indicator of potential loss of control and thus an essential precursor characteristic to prohibit.
Case against: Critics might argue that this characteristic is extremely difficult to objectively measure or prove in advance, potentially leading to subjective judgments and premature actions against AI systems that are not truly dangerous.
Estimated impact: This characteristic provides a key trigger for mandatory pauses, sequestration, and potential destruction of AI systems, even if they don't yet exhibit other superintelligent traits.
As a condition of maintaining a charter, if a charter is revoked due to a violation and that finding is upheld on appeal, the person (charter holder) forfeits their intellectual property and assets to ensure the safe and complete destruction of all systems and hardware related to the violation.
Section 12(d)(2)
Why it matters:This is a highly punitive and significant consequence, which, while explicitly stated, is nested within the operational details of the charter system rather than being a standalone prohibition.
Case for: Proponents would argue that such a severe consequence is necessary to ensure absolute compliance and to provide the government with the means to neutralize dangerous AI systems and prevent their resurgence, especially given the existential risks involved.
Case against: Opponents would contend that this provision represents an extreme government overreach, potentially violating constitutional protections against excessive fines or takings. They would argue it creates an unacceptable level of risk for AI companies and could deter all investment and innovation in the sector.
Estimated impact: This provision could lead to the complete dissolution and asset seizure of an AI company found in violation, representing an unprecedented level of government power over private enterprise in the tech sector.
Prohibits Federal funds from being used for any activities that violate sections 8, 9, or 12 or related rules, with one exception: the Department of Artificial Intelligence may use Federal funding for research relating to defensive cybersecurity measures that utilizes artificial intelligence systems that have the capacity to access secured digital or physical infrastructure without authorization (a superintelligence precursor characteristic).
Section 16
Why it matters:It's a standard funding restriction with a specific, technical exception. The exception is crucial as it allows the DAI to research certain AI capabilities that are otherwise prohibited, but only for defensive purposes.
Case for: Proponents would argue that this exception is vital for the DAI to understand and defend against the very threats it is designed to prevent. It allows for controlled research into dangerous capabilities to develop countermeasures.
Case against: Critics might argue that allowing the DAI to develop AI with 'superintelligence precursor characteristics,' even for defensive purposes, creates a dangerous loophole and could inadvertently accelerate the very risks the bill seeks to mitigate. They might question the safeguards in place to prevent misuse or unintended consequences within the Department itself.
Estimated impact: This provision creates a specific, limited pathway for the federal government to engage in research on AI capabilities that are otherwise broadly prohibited, enabling defensive innovation.