Layoffs at 27%, productivity up 9.6%! Morgan Stanley's multinational survey reveals: AI has officially rewritten corporate income statements and employee rosters.

Notice that Morgan Stanley, using an AlphaWise survey covering five countriesthe U.S., U.K., Germany, Japan, and Australiaand focused on five major industriesbanking, software and services, technology hardware, semiconductors, and professional serviceshas reached a conclusion: AI is no longer just a narrative; it has already begun to rewrite companies' employee rosters and income statements. The survey shows that over the past 12 months, surveyed companies directly cut 12% of positions because of AI, while another 15% of positions were not backfilled after employees left, meaning a combined 27% of roles were affected. The offsets were 22% new hiring, 25% employee retraining, and 16% internal transfersultimately producing an average net reduction of 5% across companies in the five countries. By country, Japan had the most severe net reduction (10%), followed by the U.K. at 6%, the U.S. at 5%, and Australia at 4%. Germany was the only country with net headcount growth (1%) and also had the highest retraining ratio. Industry divergence is equally clear. Semiconductors had the highest net reduction at 8%, ranking first among the five industries; software and services followed closely (7%); technology hardware was 5%; professional services was 4%; banking had the lightest net reduction at only 3%, but its share of retrained employees reached 26%, making it the most active industry in "cutting while training." It is worth noting that offshore roles bore the brunt: 41% of offshore employee positions were eliminated or not backfilled, and large companies with more than 10,000 employees cut offshore teams most aggressively. Junior and mid-level employees with 2 to 5 years of experience were the hardest hit by layoffs and non-backfilling, but they were also the main source of new hiring and internal transferswhat AI eliminates is not a certain type of person, but a certain type of skill. What truly concerns investors is productivity. The report shows that over the past 12 months, surveyed companies achieved an average net productivity gain of 9.6%: U.K. companies were highest (10.3%), software and services led all industries at 10.4%, and semiconductors lagged at the bottom (8.2%); small companies with fewer than 50 employees actually saw the largest gains (11.3%). By function, IT/software development and customer service/support are the two areas where AI has concentrated the greatest efficiency gains: in banking, the gains are concentrated in IT/software development and finance/FinOps, while in professional services, customer service support ranks first (52%), followed by IT/software development (49%) and legal and compliance (48%). Looking ahead to the next 12 months, all countries and all industries are placing their greatest expected efficiency gains on IT/software development71% of U.S. companies expect this, 92% in the software industry, and 86% in semiconductors. The report also contains a set of easily overlooked data: surveyed companies have on average already implemented AI solutions for 2.9 years, with U.S. companies actually having the shortest period (2.7 years); large companies with more than 500 employees and annual revenue above $5 billion adopted it earliest. This means U.S. AI penetration is still in the early-to-mid stage, and "from pilot to scaled production" remains one of the main challenges companies acknowledge, compounded by the three major obstacles of trust and security, data readiness, and legacy system integrationspending on the AI infrastructure and tools layer is far from peaking. Following the logic of its data, the beneficiary directions in U.S. equities are already quite clear: the software industry has the highest productivity gain (10.4%), 92% of software companies expect the largest efficiency gains in the coming year to remain in software development itself, and the penetration rate of AI coding and agent platforms will only continue to rise. Platform providers that embed AI capabilities into workflows, such as Microsoft Corporation (MSFT.US), ServiceNow (NOW.US), Salesforce, Inc. (CRM.US), Palantir (PLTR.US), and Adobe, are the most direct beneficiaries under this logic.