President Donald Trump pronounces remarks at the top of the White House AI in the Auditorium Andrew W. Mellon in Washington, DC, Wednesday July 23, 2025. (Official White House photo by Joyce N. Boghosian)
On July 23, US President Donald Trump revealed that his administration plan for ai. Unlike the years of unhappiness and sadness on the way robots take All jobs, the plan promises to center American workers. In addition to the export proposals for American regulations on the battery and the return of AI to the infrastructure of energy and data centers, it highlights new learning, IA-Literacy initiatives and recycling for workers. The race for an advanced AI would mean, as Trump has promised in a summit On AI and energy in Pennsylvania last month, “many jobs”.
But promising jobs are not the same as workers’ power. Experts predict that AI will not lead to certain losses, but reshaped The American economy on a scale much higher than the relocation of manufacturing in recent decades. Despite the label as centered on workers, Trump’s plan avoids difficult questions about how AI will reshape the future of work and which will have a say in this process. WHO decide If work should be created or eliminated? Who protect workers who warn uncontrollable Have? And who guarantees that the advantages of AI are shared?
Trump administration IA plan described several steps to help workers. He to propose The development of literacy and skills of AI, initiatives to support the construction of the infrastructure of the data center, and the recycling and “update” of the individuals moved by the AI. He direct Agencies such as the Ministry of Labor to identify the essential professions for the development of the data center and create new learning programs. The economic message is clear: AI arrives, and workers can help build data centers that food left behind.
But while workers are present in the AI vision of the Trump administration, they are not at the driver’s headquarters. The administration approach is accompanied by the realities of AI development, such as the fact that data centers do not create a lot Jobs once the construction is completed. For example, take the Elon Musk “colossus” data center in Memphis, fueled by gas generators who release Toxic chemicals in its surrounding community. Despite billion In investment and environmental costs, even the most generous work forecast do not compensate for losses From the factory of the domestic apparatus, he replaced.
Most importantly, the action plan mainly targets a skills difference rather than the gap in the governance of AI. Nowhere are workers invited to shape how AI is deployed at the workplace, how it is governed or how its earnings are distributed. Without protection of work and changes in negotiation power, even large -scale learning and recycling programs will not answer these fundamental questions. The proposed AI IA research hub Would find adoption, displacement and wages, the treatment of the impact of AI as something to measure, and not as something that workers can shape.
The erroneous premise of the Trump AI plan is that the economic change of AI is inevitable, and the role of the government is to help workers to follow. But it is not a vision of an AI future that we must accept. Economists like Daron Acemoglu, David Autor and Simon Johnson have noted This AI can be used to complete workers, rather than replacing them. The fact that this happens depends on political choices, not technological fate. And the historian Jason Resnikoff has supported This “AI” could well be the last chapter of a long history of societies using the language of innovation to degrade and put the line on the side, just as Ford invented the term “automation” in the 1940s to give its unions a brilliant neutral technological. The real question is therefore not to know how workers can adapt to technology, but how they can help decide what it becomes.
A program of AI truly focused on workers would treat workers not as disruption losses, but as co-creators of a new economy. Above all, this approach would seek workers to negotiate how AI is deployed, as in writers Guild of America and Sag-Aftra chords and a recent organization against the workplace monitoring. This would protect those who are expressed when AI is poorly used, like the Openai denunciators who warned against dangerous and not tested AI systems. And it would face the structural imbalances created by AI. Which could include new ideas like a token On AI results to ensure that companies cannot simply avoid labor costs by replacing human workers with not accessible machines.
This approach begins by recognizing that workers centered on workers requires a fundamental change in power. New organizational models, such as Coalition of technological workersAlready work to connect the whole AI battery, from coders to Amazon warehouse workers, in a common fight. Unions have recently formed AI working groups and started engaging declares to put railings on the AI of the workplace, including bills like “No Robo Boss Act”, which just adopted the State Senate of California. Far from the caricature “luddite” by criticismthe workforce also innovates, from Ai chatbots To help unions serve members applications which allow workers to fight against wage theft.
The federal government should also do more. Because the national labor relations law has been interpreted by the courts be ahead State regulations, rethinking negotiation rights at the federal level is essential. Like only one example, the Harvard Law work center and a fair economy propose Extend compulsory negotiation rights so that union workers cover the decisions to deploy AI to the workplace. With union membership at historical The stockings, the policy must also focus on strengthening protections for non -unionized workers in emerging sectors and in temporary employment.
As an immediate following stage, the federal government can work to extend protection against reprisals for workers in AI and technology industries, mandate Algorithmic transparency on how AI is deployed at the workplace and modernize labor law so that the workers of the platform who connect to applications have the even Protections as those who light up in the offices. Together, these policies change their power, not just skills, in the economy that AI redid.
Henry Wu is a recent graduate of the Yale Law School who is researching AI governance. He writes personally.
