Lofgren’s Science Committee plan: rebuild NIST’s AI role, make developers liable for harm and write tighter spending laws

The San José Democrat in line to chair the House Science Committee says she wants to ‘reinforce the capacity of NIST to act’ on AI and to assign liability to AI developers for harm. The White House calls Democrats’ planned investigations ‘sham’ inquiries and says its own science policy restores rigor.

Rep. Zoe Lofgren, in glasses and a dark jacket over a blue top, speaks at a podium microphone against a dark blue background

Top image: Rep. Zoe Lofgren, the San José Democrat expected to chair the House Science, Space and Technology Committee in January. Photo: Gage Skidmore / CC BY-SA 2.0 via Wikimedia Commons

Rep. Zoe Lofgren, the San José Democrat expected to lead the House Science, Space and Technology Committee when Democrats take the majority in January, says one of her first goals on artificial intelligence is to rebuild the federal agency that writes the government’s voluntary AI safety frameworks. “One of the things that we need to do is to reinforce the capacity of NIST to act,” she told Bloomberg Government this week, referring to the National Institute of Standards and Technology, which the Trump administration has targeted for budget cuts and workforce reductions.

In interviews over the past week, Lofgren has also said Congress should “assign responsibility and liability to the developers of AI for harm that is caused,” and that her committee will investigate canceled grants and withheld research funds at agencies such as the National Science Foundation. The White House rejects that framing. It says its policies restore scientific rigor, and a spokeswoman has said Democrats who “pursue sham investigations and radical left policies” are “choosing to work against the American people.”

Oversight first

Lofgren, the committee’s ranking Democrat, announced Nov. 4 that she would seek the chair after Democrats won both chambers. In a Living on Earth interview that aired the week of Nov. 13, she described two jobs: oversight “to uncover what the administration has done that has damaged the science enterprise,” and legislation. She noted that the committee has a bipartisan record, saying the bills it passed this Congress were “either unanimous or close to it.”

On the National Science Foundation, Lofgren said Congress rejected the president’s request for a $3 billion cut and funded the agency near its fiscal 2024 level, but “the money does not appear to be going out.” She said she has discussed with the incoming chair of the Appropriations Committee whether Congress should be “more prescriptive in the crafting of authorizations and appropriations, so that the opportunity for mischief is severely curtailed.” Inside Climate News reported that House Democrats see restoring federal science as a top priority.

Her staff has already begun. Committee Democrats sent preservation letters this month to agencies including NOAA and NASA, as the Chronicle reported, and Lofgren said she sent a letter to Energy Secretary Chris Wright about the department’s reorganization plans and is preparing renewed records requests to EPA Administrator Lee Zeldin.

A blue-and-white sign reading NIST, National Institute of Standards and Technology, at a campus entrance framed by trees and lawn
NIST’s campus in Gaithersburg, Md. The agency developed the government’s voluntary AI risk management framework. Photo: G. Edward Johnson / CC BY 4.0 via Wikimedia Commons

The AI agenda

NIST, a Commerce Department agency, published the AI Risk Management Framework that many companies use as a voluntary benchmark. Lofgren told Bloomberg Government that strengthening its ability to act would be a priority.

On broader AI regulation, she told Living on Earth that three House committees share most of the jurisdiction: Science, Judiciary, and Energy and Commerce. She serves on two of them. She was cool to the idea of a new select committee on AI, saying it would “take six months to get staffed up and organized.” AI “has tremendous possibilities,” she said. “It also poses tremendous challenges, and there do need to be constraints and controls.”

Lofgren has convened AI companies before. She held a roundtable in September with representatives of Hugging Face, METR, OpenAI and Anthropic, and has asked frontier AI developers to report serious incidents quickly. A chipmaking-tool export bill she has backed, the Chip EQUIP Act, is also before the Senate Commerce Committee, which scheduled a markup for Wednesday.

Official portrait of Rep. Zoe Lofgren in a magenta jacket in front of an American flag
Lofgren’s official portrait. She has represented San José in Congress since 1995. Photo: Brian Thorpe, House Creative Services (public domain) via Wikimedia Commons

The administration’s view

The administration says it has already fixed the problem Lofgren describes, from the other direction. President Donald Trump’s May 2025 executive order, “Restoring Gold Standard Science,” says the federal government contributed to a loss of public trust in science and directs agencies to make federally funded research “transparent, rigorous, and impactful” and to base decisions on “the most credible, reliable, and impartial scientific evidence available.” The order cites grant-making and guidance from the prior administration that it says was politicized.

Responding to Democrats’ preservation letters, White House spokeswoman Lauren Bis said: “If Democrats choose to pursue sham investigations and radical left policies, they are choosing to work against the American people,” according to Quartz, citing The New York Times.

The long view

For Silicon Valley, the Science Committee is not a sideshow. It writes the authorizations for NSF, NIST and the Energy Department’s national labs, the institutions that fund university research at Stanford and Santa Clara and set the technical standards that AI companies cite when they say their models are safe. A chair from San José who wants both a stronger standards agency and legal liability for AI developers would push the debate over AI rules toward enforceable obligations, at a moment when the administration favors lighter federal oversight. How far she gets will depend on how Democratic leaders divide AI jurisdiction, whether committee Republicans keep the bipartisan habits she credits them with, and whether a Republican president signs any of it. Committee Democrats post updates at democrats-science.house.gov.

Sources

Advertise with the Chronicle

Reach Silicon Valley’s decision-makers alongside reporting they trust.

The South Bay Forum

Conversation 0 comments

Join the Forum

Independent reporting on the issues that matter here, from an outsider’s perspective. Sign in to comment. An editor reads every comment before it appears.

    No comments yet. The first word is yours.