The Senate Commerce Committee voted 18-4 on May 18 to advance a five-year reauthorization of the Consumer Product Safety Commission, but a contentious amendment that would have dramatically expanded the agency’s authority over artificial intelligence-enabled consumer products was stripped from the bill after a parliamentary dispute — leaving the final product safety framework largely unchanged from its 2008 framework.
The CPSC Reauthorization at a Glance
The Consumer Product Safety Commission Reauthorization Act of 2026 (S. 2102) represents Congress’s periodic refresh of the federal agency’s statutory mandate. The CPSC, an independent federal regulatory agency, oversees the safety of more than 15,000 categories of consumer products — from household appliances and children’s toys to sporting goods and electronics. Its budget, currently funded at $145 million annually, isup for renewal every five years under the terms of the Consumer Product Safety Act.
Committee Chair Senator Maria Cantwell, D-Wash., framed the May 18 markup as a necessary update to an agency she argued had been chronically underfunded relative to its expanding mandate. “The CPSC is asked to protect American families from an ever-wider range of products, many of which didn’t exist when the last comprehensive reauthorization passed,” Cantwell said in her opening statement. “We owe them the tools, the resources, and the legal authorities to do that job.”
The bill as introduced would increase CPSC funding by 22 percent over five years, rising from $145 million in fiscal year 2026 to $177 million by FY2030. It would also authorize the agency to hire 85 additional full-time equivalent positions across its engineering, enforcement, and field inspection divisions — an increase of roughly 30 percent to the agency’s current workforce of approximately 300 staff.
The AI Provision: What It Would Have Done
The most consequential item stripped from the bill was an amendment offered by Senator Tammy Baldwin, D-Wis., that would have added a new Title IV to the reauthorization focused entirely on AI-enabled consumer products. Under the Baldwin amendment, the CPSC would have been granted authority to:
- Require manufacturers of AI-enabled consumer products — defined in the amendment as any product containing a machine learning component capable of modifying its behavior based on user interaction or environmental data — to register those systems with the agency within 90 days of commercial launch.
- Promulgate mandatory safety standards for AI systems in consumer products classified as presenting a potential risk of physical injury, including smart home devices, connected vehicles, and AI-powered exercise equipment.
- Issue recalls not just of physical products but of software updates that substantially alter an AI-enabled product’s behavior in ways that could compromise user safety.
- Establish an AI Safety Division within the CPSC, staffed by technical experts, with authority to conduct pre-market reviews of high-risk AI consumer product categories.
The amendment had support from consumer advocacy groups and a coalition of state attorneys general, but faced organized opposition from the Consumer Technology Association, the Consumer Electronics Association, and a broad coalition of technology companies ranging from established platform players to early-stage AI startups. Industry groups argued that the amendment’s pre-market review provision would create a de facto product approval regime at the CPSC — authority the agency has never held and that, in their view, Congress should not grant without extensive deliberation.
The Parliamentary Dispute That Killed the AI Title
The amendment’s removal came not through a vote against it, but through a successful point of order raised by Senator Ted Cruz, R-Texas, who argued that the Baldwin amendment constituted legislation that should originate in the House of Representatives under the Origination Clause of the U.S. Constitution. The Senate Parliamentarian ruled in Cruz’s favor, sustaining the point of order and striking the AI Title from the committee print.
Senator Baldwin expressed sharp frustration with the ruling. “The Parliamentarian has handed a victory to the tech lobby and called it constitutional law,” Baldwin said. “The American people who use AI-enabled products in their homes every day deserve better than parliamentary technicalities.” She indicated she would bring the AI safety provision as standalone legislation on the Senate floor, though the timeline for such a bill remains unclear.
“The Parliamentarian has handed a victory to the tech lobby and called it constitutional law. The American people who use AI-enabled products in their homes every day deserve better than parliamentary technicalities.” — Sen. Tammy Baldwin, D-Wis.
What Remains in the Final Committee Print
Despite the removal of the AI provision, the CPSC Reauthorization Act as advanced retains several significant provisions:
- Mandatory reporting threshold reduction: Manufacturers would be required to report product safety incidents to the CPSC within 30 days (reduced from the current 180-day window) when an incident involves a product that has caused a reportable injury or death. The current threshold for reporting has been widely criticized as allowing companies to delay disclosures while internal investigations proceed.
- Import screening expansion: The bill would require the CPSC to develop and deploy a risk-based targeting algorithm for screening imported consumer products at ports of entry, using incident data, country-of-origin patterns, and third-party testing results to prioritize inspections of higher-risk shipments.
- State preemption rollback for children’s products: The bill would allow states to impose stricter safety standards on children’s products than federal minimums — a significant departure from current law, which broadly preempts state regulation of CPSC-regulated products. The toy industry and retail groups have already signaled intent to challenge this provision on preemption grounds.
- Whistleblower protections: New statutory protections and financial incentives for industry employees who report product safety violations to the CPSC, modeled on similar programs at the SEC and EPA.
- Updated civil penalty structure: Maximum civil penalties for CPSC violations would increase from $120,000 per violation to $500,000 per violation, with a new aggregate cap of $150 million per calendar year — changes intended to restore the real deterrent value of penalties that have eroded with inflation since the penalty structure was last updated in 2008.
Path to the Floor and Key Outstanding Questions
The bill now advances to the full Senate calendar, though no floor date has been scheduled. Senate Majority Leader’s office indicated that the CPSC reauthorization will be bundled with other pending Commerce Committee items — including a broadband mapping fix and a pipeline safety reauthorization — in a legislative package to be considered after the Memorial Day recess.
The House version of the CPSC reauthorization, H.R. 3892, passed the Energy and Commerce Committee in March with significantly less aggressive funding increases and without any AI provisions — meaning that any final bill will require a conference committee to reconcile the two chambers’ differences. The tech industry’s successful removal of the AI Title in the Senate markup suggests that final legislation is unlikely to include the expanded CPSC AI authority that consumer advocates had sought — unless Baldwin’s promised standalone floor bill gains traction.
For now, the CPSC will continue regulating AI-enabled consumer products under its existing general authority under the Consumer Product Safety Act — authority that does not specifically address algorithmic behavior, software updates, or machine learning systems, and that the agency itself has acknowledged is a poor fit for the regulatory challenges posed by adaptive AI systems. Whether Congress will act to fill that gap before the next five-year reauthorization cycle remains the central unresolved question surrounding the CPSC’s regulatory future.