Democrats are racing to define their party’s position on regulating artificial intelligence in a midterm election cycle where the issue has gained traction among voters.
The party has no shortage of proposals — at least three new ones dropped on Wednesday — but no uniform position other than rejecting President Trump’s hands-off approach that lets AI companies “self-regulate.”
“Entering into a voluntary agreement that is entirely unenforceable and that allows the industry writ large to police itself is not the right response at this moment, given the potentially grave harm and danger that industry leaders themselves are now flagging for the American people,” House Democratic leader Hakeem Jeffries of New York told reporters last week.
People were interested in these podcasts
His comments followed Mr. Trump’s White House meeting with top AI executives, who signed a voluntary pact to conduct internal and external safety audits as they develop models and set up committees under their board of directors to monitor and react to those findings.
Mr. Jeffries and Senate Democratic leader Charles E. Schumer of New York have underscored the urgency of congressional action amid a rise of rogue AI agents and warnings from industry leaders that the rapidly developing technology could one day wipe out humanity if not checked.
They have not yet advocated for specific regulatory proposals, deferring to ongoing discussions among members of their caucuses about the best approach.
The House Democratic Commission on AI, led by Reps. Ted Lieu of California, Josh Gottheimer of New Jersey and Valerie Foushee of North Carolina, is developing recommendations to present to the caucus, along with proposals from “thought leaders” such as Rep. Lori Trahan of Massachusetts, Mr. Jeffries said.
Ms. Trahan on Wednesday released a discussion draft of a bill to provide legal recourse for those harmed by AI misconduct.
The Clear Liability for Artificial Intelligence Misconduct, or CLAIM, Act, would establish a federal cause of action for injured parties to bring legal claims against AI developers if their systems or rogue agents engage in deliberate or negligent harmful conduct.
“When someone breaks the law and hurts you, you can take them to court. That shouldn’t change just because the wrongdoer is an AI agent,” Ms. Trahan said in a statement. “Developers have already built systems capable of acting on their own and causing real damage. The CLAIM Act makes sure they answer for what their systems do.”
Ms. Trahan’s liability bill is designed to complement bipartisan AI safety legislation she cowrote with Rep. Jay Obernolte, California Republican.
Their FRONTIER Act would require AI developers to publish frameworks for assessing and addressing catastrophic risks and transparency reports on the capabilities and restrictions of their models, which would be subject to annual independent compliance audits.
It would also create an under-secretary position at the Department of Commerce to oversee AI security and require developers to report to that person any critical safety incidents or catastrophic risks arising from their models.
In the Senate, the Commerce Committee has been working for years on AI regulation.
Bipartisan bills that advanced through the committee last Congress when Democrats controlled the Senate never made it to the floor. Now under Republican control, AI bills have yet to even make it out of committee.
Republicans have privately blamed Sen. Maria Cantwell of Washington, the top Democrat on the Senate Commerce Committee, for stalling bipartisan progress.
Ms. Cantwell on Wednesday outlined six principles she said need to govern any AI framework: enforceable development standards, continuous testing and auditing, transparency and accountability, public-private partnerships, protections for children and workers, and global cooperation.
She offered more specific ideas under each of the overarching principles.
Ms. Cantwell wants the National Institute of Standards and Technology to establish standards to protect AI systems against catastrophic risks, such as cyberattacks, chemical, biological, radiological and nuclear threats, autonomous agents escaping secure testing environments and other loss of human control.
She said the government and independent auditors should provide constant oversight and testing post-deployment to ensure the AI systems adhere to those standards.
For transparency, developers should provide plain-language disclosures of material safety and security risks and be required to quickly report any incidents, including loss of control over autonomous agents, failures of critical safeguards and unsafe recursive self-improvement, Ms. Cantwell said.
She also wants to provide protections for industry whistleblowers to report safety or security concerns and ensure developers can be held legally liable for foreseeable harms and unlawful conduct arising from their AI systems.
Ms. Cantwell suggests the government and AI companies partner in developing defensive AI systems to detect and contain malicious activity, identify and patch cyber vulnerabilities and protect critical infrastructure from automated attacks.
Sen. Ruben Gallego, Arizona Democrat and a potential 2028 presidential candidate, also rolled out a series of AI proposals on Wednesday as part of a broader agenda to restore and protect middle-class jobs.
For example, he is proposing “a New Deal-style public service jobs program” for AI-displaced workers funded through taxes on large AI companies.
Mr. Gallego also wants to end tax deductions for AI data centers and impose a 10% excise tax on data centers’ gross receipts to fund community investments.
Contact the author
Please read our comment policy before commenting.