WASHINGTON — As Congress weighs how to respond to rapidly advancing artificial intelligence, lawmakers are confronting a basic problem: many say they lack the expertise and staff capacity to judge what rules would work — or what they might cost innovation and national competitiveness. The debate is politically divided. Some Democrats favor stronger protections for the public, while others have discussed banning so-called superintelligence. President Donald Trump and many Republicans argue that regulation could hamper American innovation and hand an advantage to China. Rep. Sam Liccardo, a California Democrat, said Congress needs to build a stronger foundation of technical knowledge to legislate effectively. He has organized briefings with prominent technology figures, including OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei and Nvidia CEO Jensen Huang, but said it has been difficult to draw colleagues to some sessions. The shortage is especially acute among congressional staff, who often help lawmakers evaluate policy. Specialists can be drawn to private-sector jobs offering far higher pay, leaving offices with limited in-house capacity to scrutinize complex proposals. Travis Moore, a former congressional aide, founded the nonprofit TechCongress in 2015 to place technology professionals in lawmakers’ and committee offices. The program has received more than 1,200 applications for about 36 fellowship positions over the past two years. Of nearly 130 fellows selected since its launch, 32 were later hired full-time, Moore said; about two-thirds have continued in public service. The program does not accept funding from corporations or technology companies, according to Moore. He said the central obstacle is not a shortage of interested specialists, but whether Congress has the will and capacity to bring their expertise into its work. Some lawmakers want a permanent congressional technology office. Rep. Mark Takano, a California Democrat, plans to reintroduce legislation to revive the Office of Technology Assessment, with bipartisan backing, his spokesperson said. The office advised Congress on emerging technology and its policy implications from 1972 until lawmakers eliminated it in 1995. Not everyone believes a lack of expertise justifies restraint. Sen. Cynthia Lummis, a Wyoming Republican who has hosted TechCongress fellows, has warned that lawmakers do not understand AI well enough to regulate it on a broad, one-size-fits-all basis. Rep. Eric Burlison, a Missouri Republican, has similarly questioned whether members should write rules for technology they barely understand. Sen. Chris Murphy, a Connecticut Democrat, called that reasoning a potential excuse for inaction. Lawmakers have had to learn about complicated subjects in other fields, he said, and he disputed the idea that only a handful of people can understand AI. Some members are trying to close the knowledge gap directly. Rep. Jay Obernolte, a California Republican, said he tutored more than two dozen House members through a bipartisan AI working group in the previous Congress. Obernolte and Rep. Don Beyer, a Virginia Democrat, have pursued master’s degrees in AI and machine learning. Rep. Valerie Foushee, a North Carolina Democrat who co-chairs a House Democratic AI policy commission, said most lawmakers have a general sense of what AI can do, but cannot confidently predict where the technology is headed. Congress has often struggled to keep pace with technological change. For decades, it largely relied on existing laws and incremental measures as the internet became part of daily life. With AI now raising questions about public safety, economic growth and the country’s ability to compete with China, the choice facing lawmakers is not simply whether to regulate, but whether they can build enough expertise to make informed decisions.