Bill Gates Renews Push For Robot Taxes As Ai Job Losses Loom Larger

Bill Gates has returned to an idea he first floated nearly a decade ago. Companies that swap human workers for machines should face the same payroll obligations those employees once carried. The Microsoft co-founder made the case again in late September during a conversation with New York Times columnist Ezra Klein. “You define a unit of labor and say, independent of whether that’s delivered by a human or robot, you are paying the same FICA tax that a human worker would pay,” Gates said, according to a Business Insider report.

FICA, the Federal Insurance Contributions Act tax, funds Social Security and Medicare. Employers and workers split the burden today. Replace the worker with software or hardware? The contribution vanishes. That tilt in the tax code, Gates argues, quietly speeds the march toward automation. But the disruption ahead looks different from past technological shifts. It could hit faster. And it may leave fewer easy landing spots for those displaced.

Gates first suggested a robot tax back in 2017. Economists pushed back then. They saw it as a levy on productivity itself. He has not dropped the thought. In an August essay on his personal site, he laid out the asymmetry plainly. Hire a person. Pay payroll taxes on their earnings. Buy a robot. Write it off as a business expense. “The tax system nudges you toward replacing people with machines,” he wrote, as quoted across outlets including TechCrunch.

A tax on AI tokens — the discrete units of computation that power large language models — or on physical robots could change that calculation. It would not stop progress. Gates does not pretend otherwise. Yet it might slow the rush. And it could generate revenue to retrain workers or strengthen safety nets. “A tax would slow the rush away from human labor a little and raise money for retraining and a stronger safety net,” he explained in the essay covered by Decrypt.

The proposal lands at a moment when lawmakers on the left have begun floating their own AI-related tax ideas. Sen. Mark Kelly, a Democrat from Arizona, introduced legislation in September that would impose excise taxes on digital advertising, AI usage, and excess profits from large AI firms. Revenue would flow into a trust fund aimed at workforce retraining, Bloomberg Tax reported on Sept. 24. Rep. Greg Casar of Texas has pushed a bill tying an AI token tax rate to the unemployment level. The higher the jobless rate, the steeper the tax.

These ideas share DNA with Gates’ thinking. They treat automation not as an unalloyed good but as a force that shifts costs onto society. Yet industry groups have rejected the notion. The International Federation of Robotics told Manufacturing Dive that Gates’ plan aims to “solve a problem that does not exist.” Automation, the group said, creates jobs through higher productivity. Taxing the tools would hurt competitiveness. That stance appeared in a Manufacturing Dive article published Sept. 25.

Nvidia chief Jensen Huang offered a different counterpoint days after Gates’ essay appeared. “I love the heck out of Bill, but I don’t see what he sees,” Huang said on Fox Business. He pointed to surging demand for electricians, plumbers, and other trades as data centers multiply. AI, in his view, remains a net job creator. The exchange, analyzed in a post on Alex Goryachev’s site, highlights the split between those who fear structural unemployment and those betting on complementary growth.

Gates himself acknowledges that AI has so far created more work than it has destroyed. But he sees a crossover coming. Accounting, legal research, telesales, customer support. These roles could vanish in the next few years once models prove reliable enough for 24-hour service in any language. Blue-collar tasks will follow. “I think ‘smart’ robots will begin to compete with people on some physical tasks — in the construction and hospitality industries, for example — by the end of the decade,” he warned in the essay referenced by CBS News.

That speed worries him. Past industrial changes unfolded across generations. Societies adapted gradually. This wave arrives in a decade or two. Workers in their 50s displaced from construction cannot simply pivot to elder care and find the switch fulfilling. The human cost runs deeper than spreadsheets suggest.

Hence his second proposal. Designate certain jobs as “human reserved.” Child care, elder care, portions of teaching and medical work. Society could decide in advance to keep these roles off-limits to full automation. Or phase AI in slowly. The ceiling, Gates has said, sits around 40 percent of all jobs at the high end. He frames it as a deliberate choice. Economic signals alone should not dictate every outcome. “Just because the economic signals say that it’d be lower cost to use an AI doesn’t mean society has to do that,” he told Klein, per the Business Insider account.

The idea echoes earlier calls for preserving human connection in care work. It also invites practical questions. Who decides the list? How does it evolve as technology improves? Gates expects the reserved category to change over time. Some roles might open to AI after a deliberate delay. Others could stay protected for social reasons.

Critics see risks. A tax on tokens or robots could raise costs for services that make health care and education more affordable. Gates has carved out that concern himself. He does not want the policy to block genuine productivity gains in those fields. Enforcement poses another hurdle. Measuring AI usage across cloud providers, custom models, and hybrid workflows is no small task. Token counts offer one proxy. But definitions shift quickly.

Still, the core insight holds appeal for some policymakers. The current code subsidizes capital at labor’s expense. A London School of Economics Business Review analysis from September spelled it out. Replace a $100,000 worker and the employer saves the 7.65 percent employer payroll contribution. The AI system arrives without that add-on. Over time, this distorts decisions and erodes the payroll tax base that supports retirement and health programs for an aging population.

Recent academic work adds weight. A new paper in Economics Letters models a tax on the installed robot stock. When redistribution effects outweigh reduced investment, the levy can raise output in the short run by boosting demand. The findings, published Oct. 1, suggest targeted automation taxes need not choke growth if designed with demand dynamics in mind.

Democrats appear divided on execution as midterms near. Some favor token taxes that scale with unemployment. Others push data-center excise taxes or equity grabs from big AI firms. Sen. Bernie Sanders has proposed a 50 percent one-time tax on equity of large AI companies to seed a sovereign wealth fund. Sen. Elizabeth Warren has pressed tech giants on deductions they receive under the 2025 tax law for AI and data-center investments. Letters sent in late September, covered by Bloomberg and others, signal continued scrutiny.

Gates frames his suggestions as conversation starters, not finished blueprints. He calls for new governance institutions at national and global levels. Preparation time is short. “The people who need the most time are the ones who have the least — the accounting worker who’s replaced by a bot or the $20-an-hour worker who loses their job to a $10-an-hour robot,” he wrote.

His foundation plans to direct $200 billion over the next two decades toward sectors likely to face pressure. That bet reflects a belief that work carries value beyond paychecks. It anchors identity. It structures days. Societies ignore that dimension at their peril.

Whether Washington acts remains uncertain. Partisan gridlock, industry lobbying, and genuine disagreement over AI’s net effects all complicate progress. Yet the conversation Gates rekindled refuses to fade. The tax code’s quiet bias toward machines has grown harder to ignore. And the pace of change has made delay riskier. Policymakers may yet find ways to balance efficiency with fairness. Or they may watch the transition play out under rules written for an earlier age. Either path carries consequences that will stretch for decades.