AI adoption is reshaping the labor structure, with banks, hardware, and semiconductors skewing toward net beneficiaries

Overview

This report is Morgan Stanley's sector deep-dive follow-up on "AI Adoption and the Future of Work," shifting the research focus from aggregate employment trends to industry- and company-level impacts. The report uses AlphaWise survey data from April 2026, covering companies that have implemented or been developing AI solutions for at least 12 months and plan to continue advancing them over the next 12 months. The core conclusion is that AI has already driven both labor structure adjustments and productivity gains, but the investment implications differ significantly across industries.

Core views

The report argues that banks benefit from lower costs, higher productivity, and improved operating leverage, with AI used more in IT, finance, customer service, and operations rather than simply replacing front-office roles. Tech hardware and semiconductors can both improve their own operating efficiency through AI and benefit from rising demand for chips and AI infrastructure. The key differentiator in software is whether companies can combine proprietary data, embedded workflows, and effective AI pricing. In professional services, long-term value comes not only from task automation or workforce reduction, but from embedding AI into core operating infrastructure and creating new products; among these, models with proprietary data and testing capabilities are better positioned, while staffing and outsourcing models face greater disruption and pricing pressure.

Analysis framework

The report combines AlphaWise primary survey data, cross-industry sample comparisons, employment and productivity metrics by country and industry, and Morgan Stanley sector analysts' assessments of operating leverage, technology budgets, data readiness, workflow embedding, and business model risks at covered companies to derive the impact of AI adoption on industries and companies.

Methodology notes

• primary_researchAlphaWise Survey

  Primary research

  An online survey conducted in April 2026 with 808 professionals familiar with their companies' AI strategy and implementation, covering the United States, the United Kingdom, Germany, Japan, and Australia. Sample companies came from the banking, software & services, tech hardware & equipment, semiconductors, and professional services industries, all of which had implemented or been developing AI solutions for at least 12 months and planned to continue advancing them over the next 12 months.

• sector_analysisSector Deep Dive

  In-depth sector comparison

  The report compares the impact of AI on jobs, productivity, business models, cost structures, and revenue opportunities across industries, identifying net beneficiary sectors and business models facing greater disruption.

• labor_productivity_analysisNet Job Impact and Productivity Gain

  Net job impact and productivity gains

  The report measures the labor and efficiency effects of AI adoption using indicators such as job eliminations, positions not backfilled, new hiring, retraining, redeployment, and productivity improvements.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Bank stocks

  Net beneficiary sector of AI adoption

  Strengths

  Improved costs, productivity, and operating leverage; large banks have advantages in data governance, technology budgets, and enterprise-scale training.

  Weaknesses

  Realization of benefits depends on system integration, data preparation, workforce upskilling, and execution capability in complex institutions.

  Comparison

  Compared with software and professional services, banks face lower risk of direct AI-driven disruption on the revenue side and reflect more efficiency and operating leverage improvement.

  Risks

  Trust, security, reputational risk, legacy system integration, data readiness, and shortages of AI skills may slow deployment.

• Tech hardware and semiconductors

  Direct beneficiaries of AI infrastructure demand

  Strengths

  They can both improve their own operating efficiency and provide the chips and infrastructure that serve as the picks and shovels of AI adoption.

  Weaknesses

  Internal workforce reductions are still at an early stage, and near-term efficiency gains may be less clear than demand-side upside.

  Comparison

  Compared with service industries, hardware and semiconductors have more direct exposure to AI capex and infrastructure demand.

  Risks

  Demand cycles, capacity, competition, technology iteration, and the pace of customer capex may affect realization of benefits.

• Software & Services

  Coexistence of AI productivity gains and business model differentiation

  Strengths

  AI has a significant impact on software development productivity, and vendors with proprietary data, embedded workflows, and AI monetization capability are more likely to benefit.

  Weaknesses

  Some companies may reduce headcount as AI improves development efficiency, and traditional seat-based or services-based pricing models may come under pressure.

  Comparison

  Compared with banks and semiconductors, the benefits in software depend more on company-level product embedding and pricing power.

  Risks

  Commoditization of AI features, pricing pressure, rising customer in-house capabilities, and organizational execution risks stemming from workforce adjustments.

• Professional services

  AI creates long-term value after being embedded into core operations, but business model differences are large

  Strengths

  AI can deliver cost savings and new product opportunities; companies with proprietary data, testing, and high value-added professional capabilities are better positioned.

  Weaknesses

  Businesses relying solely on task automation or labor input are unlikely to build durable moats.

  Comparison

  Compared with tech hardware and semiconductors, AI benefits in professional services are more indirect and depend more on transformation of operating infrastructure.

  Risks

  Staffing, outsourcing, and low-differentiation services may face greater substitution, disruption, and pricing pressure.

Key data

• Total survey sample808AlphaWise online survey in April 2026; respondents were professionals familiar with their companies' AI strategy and implementation.
• Countries coveredUnited States 160, United Kingdom 163, Germany 163, Japan 159, Australia 163Country samples cover five markets in total.
• Industry samples coveredBanks 202, Software & Services 204, Tech Hardware & Equipment 150, Semiconductors 52, Professional Services 183The sample covers five industries where the impact of AI adoption is relatively pronounced.
• Net job impact across all industries over the past 12 months-5%Formed by 12% job eliminations, 15% positions not backfilled, and 22% new hiring.
• Average productivity gain over the past 12 months+9.6%Average across all countries and industries; gains were relatively higher in the United Kingdom and the United States.
• Productivity gain in banking+10.0%The report believes AI benefits in banking are still at an early stage and may continue to expand with the extension of agentic tools.
• Productivity gain in software+10.4%The software sector faces both productivity gains and job disruption, and the degree of benefit depends on data, workflows, and monetization capability.
• Net job impact in semiconductors-8%Job disruption in semiconductors is relatively evident, but the report still believes the industry benefits overall from growing AI infrastructure demand.
• Share of S&P 500 companies mentioning AI labor impactabout 10%As of 2Q26, up from about 6% a year earlier; the share is 18% among MS AI Adopters.

Impact & implications

From an investment perspective, AI is not just a cost-cutting tool, but a long-term variable that changes industry operating leverage, talent structure, and competitive moats. Banks can unlock operating leverage through data governance, technology budgets, and workflow embedding; tech hardware and semiconductors benefit from AI infrastructure demand; software and professional services are more likely to see winner-take-more differentiation, where companies with proprietary data, deeply embedded workflows, and pricing power have a better chance of converting productivity gains into revenue and profit, while models lacking differentiated data or charging based on headcount may face pricing and demand pressure.

Risks

• AI deployment-related risks include trust, security, and reputational risk.
• Insufficient data preparation and difficulties integrating legacy systems may limit the realization of productivity gains.
• Shortages of AI skills may slow enterprise-scale rollout.
• Companies in software and professional services that lack proprietary data, embedded workflows, or pricing power may face greater competitive and pricing pressure.
• Job losses are concentrated among offshore, contract, or temporary workers and employees with 2 to 10 years of experience, which may create retraining pressure at both organizational and societal levels.
• Regulatory factors in markets such as Australia may affect the pace and breadth of AI adoption.

What to watch

• Whether companies move AI from the copilot and assistant stage toward agentic workflows.
• Whether banks can convert AI-driven productivity gains into sustained operating leverage and pre-tax profit improvement.
• Whether software companies possess proprietary data, deep workflow embedding, and effective AI pricing capability.
• Whether professional services companies can embed AI into core operating infrastructure rather than remain at isolated task automation.
• The frequency of disclosures on earnings calls regarding AI labor impact, automation, slower hiring, and decoupling between revenue and headcount.
• Constraints on adoption progress from regulation, data security, legacy system integration, and AI talent supply.