What senators heard about AI, jobs and the next policy step
Six days after a Senate hearing on artificial intelligence and jobs, lawmakers have a clearer list of questions than answers: How is AI changing work, which workers face the greatest risk, and can federal data keep up?
The Senate Health, Education, Labor and Pensions Committee’s Subcommittee on Employment and Workplace Safety held “The Impact of AI on the Workforce” on July 29, 2026, at 430 Dirksen Senate Office Building in Washington. The five witnesses were Ken Clark of EmployIndy, Carol O. Rogers of the Indiana Business Research Center, Liya Palagashvili of the Mercatus Center at George Mason University, Jonathan Liebert of the Better Business Bureau of Southern Colorado and Justin Heck of Opportunity@Work.
What the witnesses told senators
The testimony did not establish that AI is causing broad, economywide employment losses. Instead, witnesses described several changes occurring at once: some tasks are being automated, others are being augmented, productivity may increase demand for complementary work, and employers may reorganize how work is performed.
That distinction matters. A job is made up of multiple tasks, and automating some of them does not necessarily eliminate the entire occupation. The effect depends on which tasks change, what work remains and whether productivity gains create additional demand.
Palagashvili’s written testimony said some of the clearest warning signs so far involve younger workers entering highly exposed occupations. She cited studies reporting weaker job entry for workers ages 22 to 25, but the testimony also cautioned that this does not prove AI is the primary cause. Hiring declines in some exposed occupations began before ChatGPT, and broader economic conditions may also be affecting early-career workers.
The career-pathway concern
Heck focused on a less visible risk: AI could reduce or hollow out “gateway” jobs that help workers move into higher-paying roles. These jobs, including some administrative, customer-service and bookkeeping positions, often provide experience and skills that workers use to qualify for their next job.
The issue is especially important for workers without four-year degrees. Opportunity@Work describes these workers as STARs, or people Skilled Through Alternative Routes such as work experience, military service, apprenticeships or community college. If entry-level jobs disappear, or if AI removes the learning opportunities inside them, workers may lose an important route to advancement even if total employment does not fall sharply.
Why federal data are falling short
Several witnesses said the federal statistical system does not yet show enough detail about AI’s effects. Household surveys generally record whether a person is employed and what occupation they hold, but do not consistently show how AI is changing the tasks within that job. Business surveys may record whether a company uses AI, but they are not routinely linked to hiring, separations, earnings, training needs or occupational changes.
Rogers said state and local workforce planners need more employer- and task-level information to guide training and economic-development decisions. She used Indiana’s existing data infrastructure as an example of how more detailed state records could complement federal statistics, while warning that occupation-level averages are too blunt for many local decisions. Indiana is an example from the hearing, not the geographic focus of this national story.
Palagashvili also described a possible shift toward independent work. Her testimony reported that nonemployer-type business applications and solo self-employment rose faster in AI-exposed sectors and occupations. She emphasized that the findings are descriptive, not causal: they do not show that AI caused the increase. They suggest that labor-market change may appear through movement from payroll jobs to contracting or solo businesses rather than through layoffs alone.
What the proposed bill would do
The main legislative response discussed at the hearing is S.3339, the AI Workforce PREPARE Act. Congress.gov shows that Sen. Jim Banks, R-Ind., introduced the bill on December 3, 2025, and that it was read twice and referred to the Senate Health, Education, Labor and Pensions Committee. It remains listed as introduced; it has not passed the Senate, cleared Congress or become law.
If enacted, the bill would establish an Artificial Intelligence Workforce Research Hub at the Labor Department and require collaboration with the Census Bureau, the Bureau of Economic Analysis and the Bureau of Labor Statistics. The proposed hub would conduct recurring analyses and scenario planning on AI’s workforce effects.
The bill also would expand AI-related questions in federal surveys, improve occupational forecasting, collect information on AI adoption and AI-related layoffs, and support research into workers, training and employment. Those provisions are proposed text, not an operating federal program.
What the bill would not do yet
Because S.3339 remains at the introduced stage, it has not created new worker protections, guaranteed training benefits, required employers to report AI use or funded a new federal adjustment program. Any such changes would depend on further congressional action, appropriations and agency implementation.
What to watch next
The next meaningful signals are whether the HELP Committee advances S.3339, whether lawmakers seek amendments or stronger data linkages, and whether federal agencies begin publishing more useful measures of AI use, task changes, hiring, earnings, training and independent work.
For workers, the hearing’s practical message is measured rather than predictive: AI has not been shown in the testimony to be causing broad U.S. job losses, but disruption may appear first in duties, hiring pathways, training needs and the organization of work. Those changes could matter well before they show up in national employment totals.
Sources
Discover more from Interactive News
Subscribe to get the latest posts sent to your email.
