Overview

• Senator Bernie Sanders’ HELP Committee minority report projects that nearly 100 million U.S. jobs could be replaced over the next decade, with fast‑food and counter workers at 89% risk, customer service at 83%, accounting at 64%, trucking at 47% and nursing at 40%.
• OpenAI’s GPDval study benchmarked AI systems against human professionals across 44 occupations, reporting high model win rates for counter and rental clerks (81%), sales managers (79%) and shipping and receiving clerks (76%), while noting that jobs entail more than testable tasks.
• Performance differed across models, with Anthropic’s Claude Opus 4.1 and OpenAI’s GPT5‑high averaging roughly 48% win rates versus humans, compared with 12.4% for GPT‑4o in the same trials.
• Economists and analysts argue the Senate report conflates task automation with whole‑job loss and label its 100‑million figure an outlier, warning that overreaching measures such as “robot taxes” could hinder innovation.
• Recent research from Yale, Stanford and Harvard finds limited economy‑wide job displacement so far, consistent signs of greater impact on early‑career workers and growing policy proposals including a 32‑hour workweek with no loss of pay, worker ownership stakes and stronger unions.