Latest action: Read twice and referred to the Committee on Commerce, Science, and Transportation.
Referred to one or more standing committees for review, hearings, and possible markup.
Specific committees not parsed from action data.
The committee may hold hearings, amend the bill, or let it die without action.
Next: Reported
Full chamber floor — Speaker / Majority Leader controls scheduling
Creates a emergency stop button system for artificial intelligence systems.
The problem
AI systems are becoming more powerful and autonomous. There are concerns about AI systems malfunctioning or causing harm without human ability to quickly stop them.
The proposal
The bill would require certain AI systems to have an emergency shutdown mechanism that humans can activate to immediately stop the AI from operating.
The impact
AI companies would need to build safety features into their products. Users and regulators would gain ability to stop harmful AI operations quickly.
An emergency button allows people to immediately stop a malfunctioning AI system. This could prevent harm from spreading or getting worse. Quick stopping could reduce damage from AI errors.
Giving regulators and operators emergency controls makes AI systems safer for the public. It adds an important safeguard for dangerous situations. This builds confidence that AI can be controlled.
Requiring emergency buttons pushes companies to think about safety from the start. It makes safety a key requirement, not an afterthought. This could improve overall AI system design.
An emergency button helps stop AI that is being used to spread misinformation or cause harm. It gives authorities a tool to act quickly in crises. This protects public welfare.
Adding emergency buttons and safety requirements costs money and time. Companies might invest less in AI development due to new rules. This could slow innovation in the United States.
The bill does not explain exactly which AI systems need buttons or how they should work. This creates confusion about what companies must do. Unclear rules are hard to follow and enforce.
Some AI systems run across many computers and networks worldwide. Creating a single emergency button for complex AI may be technically impossible. Different AI types might need very different approaches.
Bad actors might try to abuse emergency buttons to shut down helpful AI systems. Governments might use buttons to censor or control beneficial AI. An emergency button is powerful and could be weaponized.
Generated by AI from the bill text. Not a prediction and not legal or policy advice. The percentage is the model's own rating of each item — self-reported, and not a measure of how accurate it is.
Loading map…
Loading...
0 citizens have already voted
Free · No spam · See how your representatives voted
Be the first to weigh in
Your vote starts the conversation
How citizen engagement works
Citizens vote & comment
Public opinion data is aggregated and updated in real time.
Patterns become visible
Representatives can see constituent sentiment broken down by district.
Your voice on the record
Contact your rep directly — one constituent call can shift a staffer's priority list.
Research shows that constituent contacts — especially calls — are among the most effective ways to influence a legislator's position on active bills.
Individual contributions, grouped by employer · FEC
Where their individual donors work · 2026 cycle
Money from people who work there, not from the company. Not tied to this bill.
Itemized individual contributions via FEC, grouped by the employer each contributor self-reported. Companies and unions cannot contribute to federal candidates from their treasuries, so these totals are money from people who work at an employer — never from the organization itself. They are campaign contributions covering a full election cycle and are not connected to this bill.
Senate LDA lobbying disclosures
Couldn't load lobbying disclosures right now.