Bloomberg Economics estimates 380M jobs at risk from AI disruption globally

The robots aren’t coming for everyone equally. A new Bloomberg Economics report finds that 27% of employment in advanced economies faces meaningful exposure to AI’s first-order effects, nearly triple the roughly 10% estimated for emerging markets.

The gap tells a story about economic structure. Wealthy nations have spent decades shifting toward knowledge-based work, the exact category of jobs where large language models and generative AI tools perform best.

The numbers behind the disruption

Bloomberg Economics isn’t alone in flagging the scale of potential disruption. Goldman Sachs has estimated that around 300 million jobs globally could be affected over the next decade. Bridgewater’s analysts have gone further, suggesting up to 18% of US jobs may face displacement within just five years.

Early labor market data is already showing cracks in specific corners. US occupations with high AI exposure saw employment decline by 0.2% between May 2024 and May 2025. During the same period, the broader US labor market grew by 0.8%.

The impact extends beyond existing workers. Recent graduates who majored in fields highly exposed to AI have seen a significant decline in initial earnings since ChatGPT launched in November 2022.

A mixed picture across geographies

Not every country is experiencing the same trajectory. UK labor market data offers an interesting counterpoint to the broader alarm bells. Vacancies in AI-exposed roles actually began declining before ChatGPT’s arrival, suggesting structural shifts were already underway independent of generative AI. More notably, those same UK roles showed a recovery pattern after 2024, indicating that the immediate impact of AI on British employment has been relatively contained so far.

Goldman Sachs’ 300-million-job estimate and Bridgewater’s five-year displacement forecast both assume continued rapid improvement in AI capabilities.

The Bloomberg Economics report ultimately frames AI employment risk not as a future scenario but as a present reality with measurable early indicators. The 0.2% employment decline in high-exposure US occupations while the rest of the market grew is a data point, not a prediction.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.