As Americans grow increasingly worried that AI will wipe out millions of jobs and create a permanent new underclass, tech billionaires are rushing to reassure us not to worry – the subtext being: please don’t bring out the anti-AI pitchforks.

Even Elon Musk, who recently merged SpaceX with his AI company, has joined the effort, essentially telling people “don’t worry, be happy” about AI. Musk wrote last month that “Universal HIGH INCOME via checks issued by the Federal government” would save everyone thrown out of work by AI.

Musk says that AI will create such gigantic productivity gains and such vast economic abundance that there will be no inflation and no need to “worry about squirreling money away for retirement in 10 or 20 years”. “Saving for retirement will be irrelevant,” he added.

As part of this reassurance chorus, Sam Altman’s OpenAI issued a report last month that said: “The promise of superintelligence is extraordinary … [AI] will speed up scientific and medical breakthroughs, significantly increase productivity, lower costs for families by making essential goods cheaper, and open the way for entirely new forms of work, creativity, and entrepreneurship.”

Peter Thiel, the billionaire cofounder of Palantir, has also sought to minimize worries about AI, saying, “It’s more than a nothing-burger, and it’s less than the total transformation of our society.” Thiel added that if we don’t develop AI, “the alternative is just total stagnation.”

With Americans battling to block new data centers across the US, Musk and other tech billionaires are seeking to lull the masses into believing that they shouldn’t fear AI and certainly shouldn’t fight against it. But as Musk and other billionaires maneuver to make tens of billions of dollars from AI, everyone should be doubly dubious about their assurances.

People should be worried considering that Dario Amodei, Anthropic’s CEO, said AI could eliminate half of all entry-level white-collar jobs in one to five years and lift the unemployment rate to 20%. The CEO of Microsoft AI, Mustafa Suleyman, warned that most white-collar work “will be fully automated by an AI within the next 12 to 18 months.”

As AI is rolled out across various industries, it is likely to have another important impact on workers, beyond any layoffs – AI will be used to control and surveil workers and pressure them to increase their productivity.

A Fox News poll found that nearly one-third of Americans fear that AI will eliminate their job over the next five years. But the Trump administration seems singularly unconcerned. Donald Trump has enthusiastically embraced AI and sought to speed the way for AI companies to advance and build data centers. Kevin Hassett, a top Trump economic adviser, has downplayed job concerns, saying AI isn’t costing anyone their jobs even as tech layoffs mount.

Young people entering the job market should be especially worried. They shouldn’t be comforted by Musk’s claims that AI will bring some magical economic nirvana. Many economists disagree strongly with Musk’s bullish prediction that AI will result in super gains in productivity and an overabundance of goods and services.

It would be far wiser to heed Bernie Sanders’ warnings than Musk’s super-optimism. As the top Democrat on the Senate health, education and labor committee, Sanders issued a report saying that AI could replace tens of millions of jobs and that we as a nation need to take major steps to protect against AI’s harms.

“We cannot sit back and allow a handful of billionaire Big Tech oligarchs to make decisions that will reshape our economy, our democracy and the future of humanity,” Sanders said. “The time for action is now.”

In sharp contrast, Musk suggests that if we just wait a few years, AI will bring untold prosperity. Sorry, Elon, but I side with Bernie on this. We need to take action now to protect against AI’s potentially disastrous effects: millions of layoffs, a huge increase in economic inequality, and the creation of a large, new underclass. Even Amodei has cautioned that AI may create “an unemployed or very-low-wage ‘underclass’”.

With some experts saying that AI is the most far-reaching technological development in history – Bill Gates said: “Of all the things humans have ever created, AI will change society the most” – we need to be prepared. We need to have a stronger safety net in place asap because AI is advancing so rapidly. Mass layoffs may come far faster than expected. (Although they might come more slowly and be smaller than we fear, too.) We can hope for the best, but we should plan for the worst.

Musk’s talk of high-income checks for everyone sounds wonderful, but that would require having a big-hearted, compassionate president and Congress to enact legislation to create such generous checks. Let’s not forget that Musk and other tech billionaires helped elect scores of rightwing politicians eager to shred or shrink the social safety net – last year those politicians passed legislation that makes it far harder to receive food stamps and Medicaid if you don’t have a job. So we can’t rely at all on Musk and other billionaires to win enactment of UHI, universal high-income checks. “I am skeptical about [the billionaires’ and big tech companies’] willingness to pay or incur the taxes necessary to sustain such proposals,” said Alexandria Ocasio-Cortez, a US representative who has teamed with Bernie Sanders to push for safeguards against AI.

There are several important steps that we as a nation should take to protect against AI’s harms.

We should guarantee health insurance for everyone. The US overwhelmingly ties health insurance to one’s job, and the fear that AI could displace millions of workers is a strong argument for the US to finally do what every other industrial nation has done: adopt a system of universal health insurance, perhaps through Medicare for All. That way, workers thrown to the curb by AI won’t need to panic about losing health coverage.

Wage insurance would be a smart, targeted program to help workers displaced by AI. Many of those workers will move to new jobs that pay significantly less than their old jobs, and wage insurance would be an important wage supplement, perhaps $10,000 a year. It would help offset workers’ lower wages and encourage the unemployed to search for new jobs. Beyond that, the US should make its deeply flawed unemployment insurance system more generous. Mississippi’s maximum jobless benefit is just $235 a week; in Florida, it’s just $275.

In the event that AI does eliminate millions of jobs, the US should put in place a New Deal-like Works Progress Administration that would create millions of jobs for the unemployed. Perhaps they’d repair the nation’s infrastructure or work in childcare centers. Amid predictions that AI could incinerate millions of jobs, it would be unwise to rely on the market to create enough jobs for all the workers displaced by AI.

Before AI wipes out a wide swath of jobs, we should vastly improve the nation’s less-than-stellar job training programs to prepare workers for whatever the jobs of the future will be, whether in healthcare, construction, green energy or other fields.

We should share the productivity gains from AI by legislating a 32-hour workweek at the old 40 hours’ pay – a move that should help reduce layoffs. What’s more, if AI means that our economy can thrive with the labor force working far fewer hours, then the US should finally enact a law mandating paid vacations for all workers: perhaps two weeks for new hires, three weeks after two years, and four weeks after four years. (Every worker in the 27 nations of the European Union is guaranteed four weeks’ paid vacation.)

Universal basic capital would give Americans some funds that would enable them to share in the wealth created by AI. This grant of capital – perhaps in the form of shares in a universal investment fund – could be given to every American or to all workers or just to people at birth. This, unlike universal basic income, would not only allow people to share in the profits and rising stock values, but would give people corporate voting rights. It would give the public a voice in running the nation’s corporations, and hopefully in running AI companies, too.

Seeing all the public anger at AI data centers, Sanders and Ocasio-Cortez have called for halting construction of all new data centers nationwide until Congress enacts some basic protections against AI. Sanders and AOC hope that such a moratorium would be a lever to win, for instance, a ban on AI companies releasing products that can threaten our privacy and the future of humanity.

To be sure, AI companies and many billionaires will vigorously oppose such a moratorium along with any meaningful limits on AI. That’s why we need a powerful people’s movement to fight for the strong safeguards that American workers, the American public and the world will need to protect against the vast potential dangers of AI.

• Steven Greenhouse is a journalist and author, focusing on labor and the workplace, as well as economic and legal issues