Policy

Bill Gates Warns AI Has Crossed Threat Thresholds, Proposes Token Taxes and Job Reserves

The Microsoft co-founder argues frontier models already present urgent biological, cyber, and economic risks that industry self-regulation cannot solve.

  • Artificial intelligence has already breached critical safety lines across biological capabilities, cyber warfare, and widespread labor disruption without adequate institutional guardrails, accordin…
  • Gates highlighted the weaponization potential of modern frontier models, singling out automated molecular design as an acute danger.
  • To counter the coming wave of white-collar unemployment, Gates outlined several unorthodox economic mechanisms.
Bill Gates Warns AI Has Crossed Threat Thresholds, Proposes Token Taxes and Job ReservesThe Scale Report

Artificial intelligence has already breached critical safety lines across biological capabilities, cyber warfare, and widespread labor disruption without adequate institutional guardrails, according to Microsoft co-founder Bill Gates. In a new public campaign and essay series, the 70-year-old philanthropist is urging governments to step in aggressively, warning that the traditional belief that technological shifts will organically generate enough replacement jobs no longer applies.

Gates highlighted the weaponization potential of modern frontier models, singling out automated molecular design as an acute danger. He argued that the danger of AI-enabled bioterrorism is roughly 50 times more concerning and probable than a natural pandemic, demanding mandatory surveillance on any system capable of synthesizing new molecules. To prevent unchecked proliferation, Gates called for formal biosecurity treaties, insisting Washington must directly negotiate computational monitoring standards with Beijing.

Radical Economic Safeguards

To counter the coming wave of white-collar unemployment, Gates outlined several unorthodox economic mechanisms. Among them is a "token tax"—a consumption-style levy on AI compute and usage designed to fund expanded social safety nets for displaced workers. He also revisited his long-standing proposal for robotics taxation, noting that once physical automation matches human factory or service capabilities, it could instantly displace nearly a third of the workforce.

Gates also floated the legal designation of "human-reserved jobs," which would bar artificial agents and robotics from replacing humans in designated sectors such as childcare, early education, or specific transitional trades. He suggested countries might need to implement import tariffs—analogous to the European Union's carbon border adjustments—to protect domestic industries relying on human labor from automated foreign competitors.

The former tech executive pushed back on standard macroeconomic reassurances, arguing that because AI replaces human cognition across multiple industries at minimal marginal cost, past technological transitions offer misleading comfort. Gates noted that entry-level analytical, support, and accounting jobs are already cheaper to automate, and companies are already quietly cutting back on junior hiring.

Gates was equally skeptical that top AI labs will rein themselves in voluntarily while pursuing trillions of dollars in valuation. While noting that leaders at companies like OpenAI and Google DeepMind privately harbor significant worries, he emphasized that private governance agreements have repeatedly given way to competitive commercial pressures.

Addressing his credibility, Gates conceded he is an "imperfect messenger" given past controversies, including his antitrust history and his scrutinized past association with Jeffrey Epstein. Nonetheless, he argued that his decades of championing technology make his current alarmism all the more critical, stating he intends to lobby world leaders directly on AI governance.

Why It Matters

Gates's pivot marks a notable break from the unbridled techno-optimism that defined the early digital era. When the founding architect of modern enterprise software argues that market forces and standard training programs cannot absorb the speed of algorithmic disruption, it signals that the debate over AI governance is rapidly shifting from abstract existential risk to concrete fiscal and regulatory intervention.

Reporting based on coverage from Artificial intelligence – MIT Technology Review.

The daily brief

The biggest stories in AI, venture and sports business — once a day.

One short email from The Scale Report. No spam, unsubscribe any time.

Read next