Silicon Valley and Washington Are Pitching AI Job-Loss Plans. Most Serve the Industry More Than Workers.

The Government's Answer to AI Layoffs Was a Doodle Game

In late March, the Labor Department rolled out a weeklong AI-literacy course delivered by text message, created in partnership with an unnamed AI start-up. The Atlantic enrolled and documented what it found. One message told recipients to ask a chatbot for "side hustle ideas." Another suggested brushing up on AI skills by drawing a hippo and seeing whether a bot could identify it.

That is the federal government's current answer to one of the most significant labor disruptions in modern history.

Anthropic's CEO Dario Amodei set the stakes clearly earlier this year: AI, he said, is "not a substitute for specific human jobs but rather a general labor substitute for humans." If he's right, the policy response so far is nowhere near the scale of the problem.

What's Actually Been Done

The Trump administration announced a goal to expand apprenticeship programs. Progress has been, according to The Atlantic's reporting, middling. The AI-literacy course from the Labor Department doesn't come close to retraining workers whose jobs may disappear wholesale.

Journalist Josh Tyrangiel, writing earlier this year, spoke with a broad roster of top business and political leaders and concluded that no one had a plan. The assessment stands.

The gap between the urgency of the warnings and the inadequacy of the response isn't a partisan failure. It's a bipartisan one. Both parties have spent more time talking about AI's promise or its dangers than building the infrastructure that would actually help workers land on their feet.

Sanders Proposes Government Ownership. Altman Says Maybe.

Senator Bernie Sanders introduced the AI Sovereign Wealth Fund Act, which would give the federal government a 50 percent stake in major AI companies including Anthropic and OpenAI. An appointed commission would hold voting shares to, in Sanders's words, "stop bad decisions that will reap massive job loss."

The proposal would require the tech industry to hand over trillions of dollars in stock. Coming from Sanders, who routinely targets what he calls "Big Tech oligarchs," that's expected.

What's less expected: OpenAI CEO Sam Altman reportedly requested a meeting with Sanders after the proposal was announced and expressed openness to a watered-down version of the idea. A tech billionaire finding common ground with Sanders on government ownership of AI infrastructure is a notable political development, whatever the motivation.

The Strongest Counterargument

Critics of heavy government intervention in AI have a legitimate case to make. Giving a federal commission voting shares in private AI companies introduces political incentives into decisions that should be driven by engineering and market realities. Government-managed technology sectors have a poor track record. There's also a real question of whether a 50 percent federal stake would accelerate or strangle the innovation that makes American AI competitive against China, which has its own aggressive national AI strategy.

These aren't hollow objections. The concern that Washington would slow-walk or politicize AI development while Beijing does neither warrants serious consideration.

But that concern, real as it is, doesn't make the current non-response acceptable. The choice isn't between government overreach and doing nothing. Workforce transition programs, portable benefits, retraining pipelines tied to actual hiring demand. None of those require the federal government to own a stake in Anthropic.

Who Benefits From the Current Show

The most visible "solutions" being floated right now either require minimal disruption to Silicon Valley's business model or, in the case of Altman's reported openness to a diluted Sanders plan, may actually serve the industry's interests.

Government legitimacy for AI companies, in the form of a federal ownership stake, could function as a kind of political insurance policy. If Washington owns a piece of OpenAI, Washington has less incentive to regulate it aggressively. Sanders's intent is the opposite, but the outcome of a negotiated, watered-down version could land very differently.

Meanwhile, text-message courses about hippo drawings give the appearance of action without threatening any business model at all.

What's Unresolved

The core question that neither Sanders's proposal nor the Labor Department's texts answer is a concrete one: when a specific job category is eliminated by AI automation at scale, what exactly happens to those workers, and who pays for what comes next?

That question doesn't have a legislative answer yet. Sanders's bill hasn't moved past announcement. The Trump administration's apprenticeship push hasn't produced measurable outcomes. And the companies best positioned to forecast which jobs are next, including Anthropic and OpenAI, are the same companies with the largest financial stake in making sure the transition costs land somewhere other than their balance sheets.

Whether any of the proposals now in circulation survive contact with the actual legislative process, or remain political signaling, will become clearer as Sanders's bill moves through committee—or doesn't.