CEO Proposes 1% Tax on Companies that Profit from AI

Khan Academy's CEO is warning about AI job displacement and says companies profiting from the new technology should pay to retrain displaced workers.

In a guest essay for The New York Times, Sal Khan painted an unsettling picture of AI's accelerating capabilities: A major call center in the Philippines, he wrote, recently deployed AI agents capable of replacing 80 percent of its workforce. This happened overnight in a country where call centers contribute up to 10 percent of GDP.

To help, Khan proposed a "1% solution" that would require companies benefiting from AI and automation to dedicate one percent of their profits (not revenues) to worker retraining. This could generate $10 billion annually.

To understand what this proposal means for businesses and workers alike, I turned to SmarterX and Marketing AI Institute founder and CEO Paul Roetzer on Episode 189 of The Artificial Intelligence Show.

A Tangible Path Forward

Here’s Khan’s math: About a dozen of the world's largest corporations now generate combined profits exceeding a trillion dollars annually. One percent of that would create a $10 billion annual fund, enough to build what Khan describes as "a centralized skill training platform on steroids," complete with online learning, skills verification, apprenticeships, coaching and mentorship for millions of displaced workers.

Roetzer sees merit in the proposal.

"I like the idea of a very tangible 1% tax," he says. "He is referencing specifically the world's largest corporations of combined profit over a trillion dollars. And 1% would be nothing to them."

The proposal arrives at a moment when AI’s advancements are disrupting and shrinking the workforce. Amazon has signaled more layoffs. Meta continues trimming headcount. The pattern is becoming unmistakable: AI capabilities increase, human roles decrease, profits soar.

The Backlash Is Coming

Khan frames his proposal not as charity but as corporate self-preservation: “If the public sees corporate profits skyrocketing while livelihoods evaporate,” he wrote, “backlash will follow through regulation, taxes, or outright bans on automation.”

Roetzer agrees this reckoning is inevitable.

"They're not going to be able to just benefit from reducing the human workforce in favor of profits forever without backlash," he says. "That eventually will cause political upheaval and societal revolt against these tools."

Roetzer has been thinking about this problem for some time, having proposed the concept of an "automation tax" last fall, or a mechanism to ensure companies can't simply pocket the savings from replacing human workers with AI and robotics.

So Far, Companies Lukewarm About Paying

Here's the uncomfortable reality: Roetzer has personally pitched high-level reskilling initiatives to organizations, and the reception has been less than enthusiastic.

"I would say the response has been pretty lukewarm at best," he says. "I think it was maybe just too abstract or people didn't want to admit the amount of disruption the technology was going to have, and so they just weren't ready to have these conversations."

But that reluctance may be shifting. Some states have already moved in this direction. For instance, Ohio's Tech Cred program provides $40 million in reimbursements for companies investing in technology education with at least eight other states running similar initiatives. Canada has comparable programs.

"I think these tech companies are going to wake up this year to the fact that they need to be putting hundreds of millions of dollars behind this," says Roetzer.

Retraining for What Future Jobs?

Khan's essay points to where displaced workers might go: healthcare (nearly two million job openings projected annually), education (a global shortage of 44 million teachers by 2030), construction (500,000 additional workers needed each year), trades (electricians and plumbers, especially) plus expanding roles in hospitality and elder care.

But Roetzer raises a harder question, one that Khan's optimistic projections don't fully answer.

"What jobs are they training for?" Roetzer asks. "What are they providing $10 billion to prepare them for when no one seems to know what the jobs three to five years from now look like?"

It's a genuine dilemma. If the technology being developed is explicitly designed to handle cognitive labor at scale, traditional retraining assumptions may not hold. The jobs of 2028 might not look like the jobs of 2025.

Despite the uncertainties, Roetzer welcomes Khan’s proposal. As the CEO of an organization that partnered early with OpenAI to build AI-powered learning into its platform, Khan is speaking from experience. He understands the transformation AI brings.

"I'm glad Sal Khan is saying these things and proposing ideas," says Roetzer. "I think we have to have more of this kind of discussion."