Search results for O*NET
AI & Enterprise
AI benchmarks miss 90 percent of human work, leaving blind spot in performance assessments
Existing benchmarks used to assess AI job performance cover only a small slice of real workplace tasks, an analysis found, making it difficult to conclude broad job capability from high test scores. U.S. AI talent and evaluation firm Mercor said about 90 percent of U.S. work hours are tied to tasks not adequately assessed. Using the Labor Department’s O*NET database, it mapped 29 benchmarks to 18,797 tasks and found only 4.9 percent met its coverage threshold.
AI & Enterprise
Stanford says AI hits entry-level jobs first, widening youth employment gap
Stanford University research found employment declines are widening among entry-level workers in jobs with high AI exposure. In an August 2026 report, economists said employment for workers aged 22 to 25 fell more in high AI-exposure occupations than in low-exposure ones. The gap for the most exposed group reached 19 percent, up from 13 percent a year earlier. The study linked the deterioration mainly to slower hiring rather than rising layoffs.
AI & Enterprise
AI really taking jobs? Anthropic says not yet
Anthropic on March 5 released a report on AI’s impact on the labour market, proposing a new “Observed Exposure” index that incorporates real usage data. It found large gaps between AI’s theoretical capability and actual use, with computer and administrative jobs among the most exposed. Higher exposure correlates with lower projected employment growth, but unemployment increases since ChatGPT’s launch were not statistically significant. It flagged weaker hiring for highly exposed jobs among 22 to 25-year-olds.