Tax break for AI training: employers gain, workers' gains unclear

What it does

This bill creates a federal tax credit allowing businesses to deduct 30% of their AI training costs for employees, up to $2,500 per employee per year. The credit covers enrollment in accredited AI programs, employee wages during training, and in-house AI training development. The Treasury, Labor, and Commerce departments must run a public awareness campaign and report annually on uptake.

Why we flagged it

The bill's core mechanism is a tax credit—a direct subsidy to employers. While framed as workforce development, it functions as a business tax break with no enforcement mechanism ensuring workers actually benefit or that training quality meets labor-market needs.

What the text implies

• The $2,500-per-employee cap may incentivize only shallow, low-cost training programs rather than comprehensive skill development, since employers can claim the credit regardless of training depth or job-placement outcomes.
• No requirement that training lead to wage increases, promotions, or job retention—employers can claim the credit for training that does not materially improve worker outcomes.
• The credit is available to all employers regardless of size or profitability, meaning large tech firms with existing training budgets may simply relabel existing programs to claim the credit (deadweight loss).
• Public outreach is mandated but no enforcement mechanism exists to verify that workers are actually informed of or able to access training opportunities; the campaign may reach employers but not workers.
• The bill does not define 'accredited' AI training programs, leaving ambiguity about what qualifies and potentially allowing low-quality or vendor-driven training to claim the credit.

Who it affects

Workers gain access to subsidized AI training that may improve job prospects and wage potential in a growing sector. However, the primary beneficiary is employers (who receive the tax credit), and the benefit is capped and conditional on employer participation—workers at firms that don't invest in training see no direct gain. The public cost (foregone tax revenue) is unquantified.

Who stands to gain

• technology companies (especially those with large payrolls and existing training infrastructure)
• AI training vendors and online education platforms
• large employers in any sector seeking to upskill workforces

Named in the bill

Internal Revenue Code Section 45BB, Treasury Department, Department of Labor, Department of Commerce, Small Business Development Centers, Trade associations, Workforce boards, Accredited AI training programs

Where it stands

4 cosponsors: 3 Democrats, 1 Republicans.

• Feb 13, 2026 — Introduced · Congress.gov: “Introduced in House”
• Feb 13, 2026 — Referred to House Committee on Ways and Means · Congress.gov: “Referred to the House Committee on Ways and Means”

Dates and quoted wording are Congress.gov's action record; the timeline shows status changes, not every procedural step.

Money around this bill

5 lobbying clients named this bill on 5 disclosure filings across 1 quarter, Jun 2026 to Jun 2026. Those filings disclosed $4,580,000 in lobbying spend. A filing names 10 bills on average, so that figure is what each filing reported, not a share belonging to this bill.

More lobbying clients named this bill than 78% of bills with at least one filing.

Josh Gottheimer, the sponsor, reported $1,000,800 in PAC receipts in the 2026 cycle.

• Google Client Services LLC — $3,570,000 on 1 filing
• Lenovo (united States) Inc. — $570,000 on 1 filing
• Cf Industries Inc — $410,000 on 1 filing
• Association of California Egg Farmers — $20,000 on 1 filing
• Alaska Power Association — $10,000 on 1 filing

Lobbying Disclosure Act filings through Jul 20, 2026. A filing shows who paid to lobby on a bill it names, not what changed.