Global AI Employment Impact Shifts to Negative

S&P Global’s latest assessment of AI’s impact on employment marks a significant reversal. The Purchasing Managers’ Index special survey shows a global net employment impact of -5 percentage points over the past 12 months from AI adoption—calculated as the percentage of businesses increasing workforce due to AI minus those decreasing.

This represents a stark shift from S&P Global’s prior report, which characterised AI’s employment effect as neutral to slightly positive. The 2026 analysis reverses that assessment to modestly negative, and the outlook is expected to worsen. The S&P Global PMI survey forecasts a further net employment impact of -2 percentage points from AI adoption in the coming year.

European Divergence: Germany and UK Report Job Losses, Italy Gains

The employment impact varies significantly across Europe. In Germany, the net AI employment impact was -2 points, with administrative and marketing roles cut while AI developers and implementation managers were added. The UK reported a more severe picture, with a net negative workforce impact of -6 points, particularly affecting administration, customer service and finance roles.

Italy stands apart with a net AI employment balance of +9 points, reflecting more firms adding jobs in marketing, graphic design, analyst and IT roles.

Delivery challenges compound the employment picture. Only 46% of AI initiatives launched in the past year are estimated to be on track to achieve positive ROI within 12 months—a figure even lower in France (43%) and Germany (38%). Just 37% of AI initiatives over the past 12 months were classified as live and delivering value, with many projects stuck in development or partial deployment.

Large Enterprises Forecast Steeper Job Cuts

Large companies (10,000+ employees) forecast a net negative employment impact of -13 points from AI in 2026, while small firms forecast +3 points and medium-sized firms +2 points. Among large enterprises surveyed, the proportion reporting AI-related job reductions was 8 points higher than the share reporting gains.

Of the S&P Global 1200 index, 994 participants (83%) had a lower head count in January 2026 compared to January 2025; only 153 (13%) experienced an increase.

Enterprise Priorities: Efficiency Over Headcount Reduction

Among enterprise AI objectives surveyed, process efficiency is cited by 64% of respondents and employee productivity by 59%, while head count reduction is cited by only 24%. This suggests that while job losses are materialising, they remain a secondary objective for most organisations.

However, automation ambitions are relatively modest. Only 22% of AI projects across surveyed organisations target a fully autonomous end state where AI operates without human intervention.

Adoption Rates and Use Cases

Across 38 AI use cases surveyed, the average current adoption rate is 50% and the average planned adoption rate over the next year is 37%. Summarisation (71%), translation (62%) and data management (61%) are the most widely adopted AI use cases.

Security-focused applications also show significant traction: 51% of surveyed organisations reported investing in AI for identity verification and access assurance, with 29% targeting full automation of that use case.

Skills Gaps and Trust Deficits Hamper Progress

Cybersecurity skills gaps had a moderate or severe impact on AI initiatives for 64% of respondents; machine learning and AI development gaps affected 59%; software development and engineering 58%; and data management and governance 57%.

Concerns about data privacy and security are cited as limitations of generative AI models by 51% of respondents; response accuracy and quality by 46%; and data quality by 38%.

Trust in third-party AI models has deteriorated sharply. In 2026, only 16% of surveyed organisations say they completely trust third-party AI models, and 30% mostly trust them, down from 24% and 42% respectively in 2023.

Only 44% of organisations with 10,000+ employees report a ‘clear, documented AI strategy’ aligned with core business goals, including dedicated roles and career paths for specialised AI professionals.

Labour Market Data Remains Inconclusive

While enterprise surveys show net job losses, real-world labour market data presents a more ambiguous picture. The Budget Lab at Yale (updated September 15, 2026) finds that the occupational mix is not yet changing in ways that clearly align with the introduction of AI into the workforce. Measures of AI usage show no statistically detectable connection to changes in employment or unemployment.

A synthetic differences-in-differences analysis of AI exposure conducted by the Budget Lab does not yet clearly indicate an AI-related labour market footprint.

Source: S&P Global