Worker Protection Gaps in Trump's AI Action Plan
On Labor Day, it's worth examining what the administration's "worker-first" AI policy actually offers displaced workers, and what it doesn't.
Last week I wrote about Trump's AI Action Plan and the concerning provision about studying Chinese AI methods under the banner of "protecting free speech." Today, since it's Labor Day, seems like the perfect time to dig into another part of the plan: worker protections.
The administration promises a "worker-first AI agenda," but when you look at the actual policies, the protections for AI-displaced workers seem woefully inadequate for the scale of disruption coming our way.
The AI Action Plan acknowledges that AI will "transform how work gets done across all industries and occupations" (White House, 2025, p. 6). Think about that for a moment. We're talking about potential disruption to millions of jobs across sectors, from customer service and data entry to legal research and medical diagnosis (World Economic Forum, 2025). This isn't like previous technological changes that affected specific industries. This is broader.
Yet the plan's main response? Retraining programs and tax incentives for AI education.
What's Actually on the Table
Let me break down what the plan actually offers workers:
The Plan Provides:
AI skill development integrated into existing workforce programs
Tax-free reimbursement for AI training under Section 132 of the tax code
"Rapid retraining for individuals impacted by AI-related job displacement"
An AI Workforce Research Hub to study the impacts (because apparently we need to study what's already happening)
Career and technical education focused on AI infrastructure jobs
What's Conspicuously Missing:
Universal Basic Income pilots to help workers during transition periods
Portable benefits that move with workers between gigs and jobs
Income support beyond just retraining (what do you live on while learning new skills?)
Specific protections for gig workers and entry-level professionals, who are most vulnerable to AI displacement
Realistic timelines for how fast retraining programs will scale up versus how fast jobs disappear
The assumption seems to be that displaced workers can seamlessly transition to new careers while maintaining their mortgage payments and feeding their families. That's not how economic disruption works in the real world.
The Infrastructure Jobs Mirage
Here's where the plan gets particularly frustrating. It heavily emphasizes creating "high-paying jobs for American workers" in AI infrastructure, such as data centers, energy systems and semiconductor manufacturing (White House, 2025).
But there's a fundamental problem here:
AI eliminates jobs broadly across service sectors from knowledge work to entry-level positions
AI infrastructure creates jobs narrowly in specialized technical fields
A displaced customer service representative can't easily retrain to become an HVAC technician for data centers
The geographic mismatch alone creates problems. Many displaced workers can't or won't relocate for new careers, especially when they have community ties, family obligations, or underwater mortgages.
We've Seen This Before
This isn't the first time we've heard promises about retraining programs solving technological displacement. We saw it during deindustrialization in the Rust Belt, and the results weren't pretty.
During the factory automation wave of the 1980s-2000s, we witnessed:
Rapid technological change that eliminated both manual and cognitive work
Skilled workers affected, not just unskilled labor
Promised retraining programs that largely failed to deliver
Geographic concentration of job losses that devastated entire communities
What did we learn from that experience?
Retraining doesn't work for everyone: Older workers particularly struggled to learn new skills
Wage replacement is rare: Most displaced workers found lower-paying work
Community effects compound individual problems: When major employers disappear, entire local ecosystems collapse
Political backlash is inevitable: Economic displacement led to populist movements and trade policy reversals
The AI Action Plan seems to be following the same playbook that failed during deindustrialization, just with updated terminology around AI and the digital economy.
How Other Countries Are Handling This
No single country has definitively effective and comprehensive policies to protect workers from all the potential risks of AI. However, the most effective efforts are coming from the European Union, followed by specific initiatives in countries like Canada and Japan. These policies focus on transparency, accountability, and preventing discriminatory practices.
European Union: The EU's AI Act takes a comprehensive risk-based approach, with strict rules for "high-risk" employment AI systems. It bans harmful practices like workplace social scoring, requires transparency and bias assessments for hiring and performance management AI, and mandates that workers be informed before deployment with rights to file complaints (Cabrera & Maier, 2025).
Canada: While federal AI legislation is delayed, provincial laws are emerging. Ontario will require AI disclosure in hiring by 2026, Quebec mandates human review rights for automated employment decisions, and unionized workplaces can negotiate before AI implementation. Privacy laws require worker notification about AI data collection (Government of Canada, 2022).
Japan: Emphasizes "agile governance" through voluntary guidelines prioritizing innovation while protecting rights. The 2025 AI Promotion Act promotes development while requiring human rights respect, and the Social Principles of Human-Centered AI mandate fairness, accountability, and transparency aligned with OECD principles (Nayak, 2025).
The contrast with the American approach is striking. While other countries are implementing specific worker protections and transparency requirements, our AI Action Plan focuses primarily on removing regulatory barriers and speeding up AI deployment.
The Dangerous Feedback Loop
Here's what really concerns me about the inadequate worker protections. If AI displacement happens faster than our retraining programs can handle, we're setting up a dangerous cycle of:
Mass unemployment in affected sectors
Political backlash against AI adoption
Pressure to slow AI development
Loss of competitive advantage the plan is trying to achieve
In other words, the plan's failure to protect workers could undermine its own goals for AI "dominance".
What Adequate Protection Might Actually Look Like
If the administration were serious about "worker-first" AI policy, here's what we might see:
Advance warning systems: Requirements for companies to provide 6-12 months notice before AI implementations that eliminate jobs, giving workers time to prepare.
Adjustment assistance: Direct income support during transition periods, not just training opportunities you can't afford to take.
Job guarantee programs: Government commitment to employment during retraining periods, removing the financial anxiety that makes learning new skills nearly impossible.
Community transition funds: Support for entire regions affected by AI displacement, recognizing that job losses create ripple effects throughout local economies.
Collective bargaining rights: Stronger worker voice in how and when AI gets implemented in their workplaces.
The Political Reality Check
The truth is, the plan's "worker-first" rhetoric seems designed to deflect criticism while prioritizing corporate AI adoption. Real worker protection would require significant government spending and business regulation, both of which directly conflict with the plan's core commitment to eliminating regulatory barriers and maximizing AI deployment speed.
You can't have a genuine "worker-first" AI policy that's also dedicated to removing obstacles for companies wanting to implement job-displacing technology as quickly as possible. The plan tries to have it both ways, but basic economics tells us who's going to bear the costs of that contradiction.
The administration is essentially asking workers to trust that the market will sort everything out, despite decades of evidence showing that technological transitions require active government intervention to prevent widespread economic hardship.
The Bottom Line
On this Labor Day, it's worth remembering that technology isn't destiny. How we manage AI's impact on work is a choice, not an inevitable outcome. The AI Action Plan's worker protections aren't just inadequate, they're a missed opportunity to show that American innovation can coexist with genuine concern for working families. Instead, we get recycled promises from previous failed transitions, dressed up with new buzzwords about the AI economy.
Workers deserve better than retraining programs that might reach a fraction of displaced people while everyone else falls through the cracks. They deserve a plan that acknowledges the scale of what's coming and responds accordingly.
Until we get serious about worker protection, all the talk about "worker-first" AI policy is just that—talk.
What You Can Do
Don't want to just read about this problem? Here are concrete steps you can take:
Contact your representatives and ask them to explain how retraining programs will match the scale of AI displacement
Research which jobs in your field are most vulnerable to AI and start building complementary skills
Attend a local town hall or city council meeting and ask about AI transition planning in your community
If you're in a workplace where AI might be implemented, start conversations about advance notice and retraining
Support organizations that are researching and advocating for comprehensive worker protection policies
The goal isn't to stop AI development, rather it's to ensure the benefits don't just flow to tech companies while workers bear all the costs of transition.
This post was written by me, with editing support from AI tools, because even writers appreciate a sidekick.
References
Government of Canada. (2022, June 16). New laws to strengthen Canadians’ privacy protection and trust in the digital economy. [Press Release]. https://www.canada.ca/en/innovation-science-economic-development/news/2022/06/new-laws-to-strengthen-canadians-privacy-protection-and-trust-in-the-digital-economy.html
Cabrera, L.L. and Maier, M. (2025, April 14). EU AI act brief – Pt. 4, AI at work. Center for Democracy & Technology. https://cdt.org/insights/eu-ai-act-brief-pt-4-ai-at-work/#:~:text=The%20EU's%20AI%20Act%2C%20the,on%20the%20protection%20of%20workers.
Nayak, R. (2025, August 13). Japan's AI legislation: Agile governance creating competitive advantages. Diligent. https://www.diligent.com/resources/blog/japan-ai-regulations
White House. (2025, July). America's AI action plan. https://www.whitehouse.gov/wp-content/uploads/2025/07/Americas-AI-Action-Plan.pdf
World Economic Forum. (2025, January). Future of jobs report 2025. https://reports.weforum.org/docs/WEF_Future_of_Jobs_Report_2025.pdf