House subcommittee hears how AI could reshape U.S. manufacturing, amid concerns about federal uncertainty

Chairman Bill Bilirakis, chair of the House Energy and Commerce Subcommittee on Commerce, Manufacturing, and Trade, opened the Feb. 12 hearing saying he was “excited to kick this Congress off with an educational hearing examining the state of American manufacturing and how it can be revolutionized with the use of artificial intelligence.” The panel heard testimony from industry and academic witnesses on how AI is being used now in factories and supply chains, and on what federal policy could do to accelerate adoption while managing risks.

The hearing’s witnesses — Jason Oxman, president and CEO of the Information Technology Industry Council (ITI); Elizabeth Reynolds, professor of the practice at MIT; Jeff Kinder, executive vice president for product development and manufacturing solutions at Autodesk; and Barbara Humpton, president and CEO of Siemens USA — described a set of use cases that subcommittee members said could help U.S. manufacturers lower costs, reduce downtime, and improve worker safety.

"AI means increased efficiency, productivity, safety, and innovation," Jason Oxman told the committee, and he urged policies that support tax incentives, workforce training, broader access to federal datasets and public‑private partnerships to speed AI deployment in manufacturing. Elizabeth Reynolds said the United States is “in the midst of the next industrial revolution” and urged targeted help for small and medium‑sized enterprises, workforce reskilling, and more financing vehicles to help manufacturing start‑ups scale.

Witnesses gave practical examples. Oxman cited an Ericsson 5G “smart factory” that used predictive maintenance and automation to improve output per employee by 20 percent and reduce manual material handling by 65 percent. Humpton described a Siemens predictive‑maintenance model that flagged potential machine failures up to six months ahead of time and a Siemens facility in Fort Worth that used a factory “digital twin” to design operations and create hundreds of jobs. Kinder described Autodesk work that used AI to reduce a NASCAR brake pedal’s weight by 32 percent and shorten the time it takes to generate computer numerical control (CNC) machining code.

Members pressed witnesses about the non‑technical issues that influence adoption. Representatives from both parties asked how to ensure cybersecurity and protect sensitive data when production systems are connected; Kinder summarized Autodesk’s approach as “build secure, run secure, and stay secure.” Witnesses and members also discussed energy: several speakers said data‑center power needs for large AI models require permitting reform and an “all‑of‑the‑above” energy strategy so computing capacity can be deployed without destabilizing local grids.

Democratic members raised broader policy concerns that they say threaten current momentum. Ranking Member Jan Schakowsky and other Democrats warned that recent administration actions — including pauses in grant and loan disbursements and disputed data access arrangements involving private sector actors — have injected uncertainty into projects that depend on federal supports such as the CHIPS and Science Act and the Inflation Reduction Act. “Uncertainty is bad for business,” Elizabeth Reynolds told the committee, adding that many announced projects are still in pre‑construction stages and rely on predictable federal obligations.

Republican members pressed witnesses on regulatory approaches and international competition. Several Republicans said the U.S. must avoid an overly restrictive regulatory path and instead emphasize standards, public‑private partnerships and incentives to keep investment in the U.S. Several members also said the United States must secure the energy and permitting pathways needed for data centers that host AI compute.

Members from both parties returned repeatedly to workforce concerns. Reynolds and other witnesses recommended scaling apprenticeship and pre‑apprenticeship programs, expanding community‑college partnerships, and increasing recruitment of women into manufacturing roles. Reynolds said the U.S. could face a shortage of nearly 2,000,000 manufacturing workers by 2033 if training and hiring do not accelerate.

The hearing produced no formal votes or committee actions. Members were invited to submit additional questions for the record; the chairman closed the hearing after asking for further written submissions by Feb. 26.

Ending: The hearing collected both cross‑party recommendations — tax and financing incentives, workforce training, energy and permitting fixes, and cybersecurity safeguards — and highlighted a central tension: industry and many members want rapid adoption of industrial AI to restore U.S. competitiveness, while some members warned that recent federal policy moves and regulatory choices could slow or reroute planned investments. The subcommittee left the topic open for additional oversight and follow‑up questions for the witnesses.