AI is more likely to boost productivity than to immediately cause a macro-level employment cliff

Overview

This edition of Sunday Start focuses on AI, productivity, and the labor market. The report defines AI as the sixth major wave of innovation since the Industrial Revolution and argues that the probability of AI significantly improving productivity over the long term is high, but its pace of development and diffusion is far faster than historical innovation cycles, so the business-cycle impact remains uncertain. Current macro data more strongly support an explanation of productivity gains driven by output expansion rather than one in which AI has already caused large-scale job substitution.

Core views

The core view is that AI will reduce the cost of specific tasks and change the composition of work, but whether it will eliminate entire occupations remains to be seen. Morgan Stanley is inclined to believe that the same level of employment can generate more output, rather than achieving the same output merely through reducing labor. Current indicators such as employment, unemployment, job openings, and quits do not show systemic weakness in AI-high-exposure industries; productivity gains are mainly coming from output growth rather than reduced hours worked. The future risk is that AI adoption could proceed too quickly, causing job destruction to outpace job creation, but aggregate demand feedback, consumption, wealth effects, the creation of new tasks, and monetary and fiscal policy can cushion the transition shock.

Analysis framework

The report uses a framework combining historical analogies, current macro labor-market data, industry AI exposure and productivity performance, the pace of corporate adoption, and policy reaction functions. It first places AI in the historical context of the previous five waves of innovation, then examines employment and productivity data in high-AI-exposure industries, and finally assesses whether demand feedback, monetary policy, and fiscal policy can smooth income and employment gaps during the transition.

Methodology notes

• Historical comparisonSix-wave innovation framework

  Places AI alongside mechanization, electrification, mass production, automation, and the IT revolution as the sixth major wave of innovation.

  This framework is used to show that each technological wave has raised concerns about job displacement, but over the long term has usually expanded economic output through productivity gains, task reorganization, and the creation of new demand.

• Labor market monitoringComparison of AI-exposed industries

  Compares the performance of high- and low-AI-exposure industries in employment, unemployment, job openings, and quits.

  The report notes that these indicators do not yet show systemic weakness in high-AI-exposure industries, and that the rise in unemployment among younger workers is only slightly above historical cycle patterns after adjusting for the overall hiring slowdown.

• Productivity decompositionDistinguishing output growth from changes in hours worked

  Determines whether rising labor productivity comes from faster output growth or from reduced hours worked and labor substitution.

  The report emphasizes that productivity gains in high-AI-exposure industries are mainly driven by output growth rather than declining hours worked, and this distinction is crucial for judging whether AI is causing large-scale substitution.

• Macro policy feedbackDemand and policy buffering mechanism

  Productivity-driven income gains, wealth effects, new job creation, monetary easing, and fiscal stabilizers jointly cushion the employment transition.

  If AI causes a temporary slowdown in employment and downward pressure on inflation, monetary policy may stimulate the economy back toward full employment; if monetary policy is constrained, automatic stabilizers and discretionary fiscal policy can partially smooth income gaps.

Asset mapping & comparison

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

• AI and data center infrastructure

  The key physical constraint and capex vehicle for AI diffusion

  Strengths

  Related capital expenditure is expected to exceed USD 3 trillion in 2025-2028, long-term productivity upside is substantial, and companies are beginning to report more quantifiable gains from AI adoption.

  Weaknesses

  Only about one quarter of capital expenditure has been deployed, infrastructure construction is still incomplete, and there is uncertainty around the pace of adoption and realization of returns.

  Comparison

  Compared with past innovation waves, AI is diffusing faster and may compress the economy’s adjustment period.

  Risks

  Capex moving too fast, delayed return realization, power and supply-chain constraints, and regulatory and policy uncertainty.

• Global labor market

  The main macro transmission channel of the AI productivity shock

  Strengths

  Current data on employment, unemployment, job openings, and quits do not yet show systemic weakness in AI-high-exposure industries.

  Weaknesses

  Some job losses and hiring slowdowns are inevitable, and unemployment among younger workers remains a market concern.

  Comparison

  Historical innovation waves ultimately tended to expand labor demand, but AI is moving faster, so adjustment pressure may be more concentrated.

  Risks

  Job destruction outpacing job creation, worsening income distribution, and short-term unemployment rising in a near-recession-like manner.

• U.S. rates and duration

  If AI or oil-price shocks lead to slower employment and disinflation, policy paths and duration assets will be affected

  Strengths

  The report notes that rates strategists favor duration, and lower mortgage rates may support a rebound in home prices in the second half of the year.

  Weaknesses

  Core PCE is still expected to rise to 3.11% year over year, and inflation stickiness limits room for rate cuts.

  Comparison

  Relative to growth risks, the FOMC still needs to balance inflation and uncertainty.

  Risks

  Higher-than-expected inflation, oil-price shocks, greater-than-expected policy patience, and rising term premium.

• Global central bank policy

  A stabilizer for employment and inflation shocks during the AI transition

  Strengths

  The Federal Reserve, ECB, BOJ, Bank of Canada, and others are expected to maintain a cautious stance, and policy can still respond to demand shortfalls.

  Weaknesses

  If the shock both raises inflation and suppresses growth, the policy tradeoff becomes more difficult.

  Comparison

  Monetary policy is the first line of cushioning, while fiscal policy provides support when monetary policy is constrained.

  Risks

  Unanchored inflation expectations, insufficient fiscal space, and delayed policy responses.

Key data

• AI-related data center and infrastructure capital expenditureExpected to exceed USD 3 trillion in 2025-2028, with only about one quarter currently deployedThe report argues that AI adoption still depends on complementary physical infrastructure that remains under construction.
• U.S. home pricesCase-Shiller national index up 0.9% year over year in January, versus 4.2% a year earlier; expected to slow to 0.5% in FebruaryRates strategists favor duration, and lower mortgage rates may help reaccelerate home prices to around 2% in the second half of the year.
• U.S. 1Q26 GDPExpected at 2.4% quarter-over-quarter annualizedCore PCE prices are expected at 4.1% quarter-over-quarter annualized, and headline PCE at 4.3%.
• U.S. core PCEExpected at 0.22% month over month and 3.11% year over year in MarchIf the BEA uses PPI legal services rather than CPI legal services, core PCE month over month could be around 0.26%.
• Federal Reserve policyExpected to keep rates unchangedThe FOMC may retain an easing bias, but emphasize that rising uncertainty means policy needs to remain patient.
• European Central Bank policyExpected to keep rates unchanged at the April meetingThe Governing Council may keep all options open and avoid giving strong guidance on the future rate path.
• Bank of Japan policyExpected to keep rates unchangedThe April meeting may lack sufficient information to assess the persistence of the shock.
• China manufacturing PMIExpected to rise to 50.6, from 50.4 in MarchThis is one of the macro items to watch this week.

Impact & implications

For asset allocation, the implication of the report is not simply to bet on AI causing an employment collapse, but to focus on the pace of AI diffusion, infrastructure buildout, actual corporate adoption, and policy responses. If AI productivity gains are released gradually and policy cushioning is effective, the macro outcome is more likely to be an improvement in medium- to long-term potential growth; if adoption far outpaces the economy’s absorptive capacity, then short-term unemployment, income distribution pressure, and policy response challenges will rise. For rates markets, if employment weakens temporarily and brings a disinflationary shock, monetary policy may tilt more toward supporting growth; for equities and the AI infrastructure chain, the pace of capex deployment and quantifiable productivity gains are the core variables for validation.

Risks

• AI adoption proceeds too quickly, causing job destruction to temporarily outpace job creation.
• Productivity gains are released rapidly but distributed unevenly, preventing some social groups from benefiting equally.
• Corporate AI capital expenditure and infrastructure buildout progress fall short of expectations, dragging on adoption and the realization of benefits.
• Oil-price and geopolitical shocks push inflation higher, making it harder for central banks to cushion employment pressure with easier policy.
• Macro data are still at an early stage, and labor-market effects may emerge with a lag in the future.

What to watch

• The pace of AI diffusion and corporate adoption, especially whether companies report more quantifiable productivity gains.
• Changes in employment, hours worked, job openings, and quits in AI-high-exposure industries relative to low-exposure industries.
• Whether labor productivity gains are driven by output growth or by cuts in hours worked and job substitution.
• The pace of deployment of data center and related infrastructure capital expenditure.
• How the Federal Reserve, ECB, BOJ, and other central banks balance growth, employment, and inflation in policy decisions.
• Key U.S. macro data such as PCE, GDP, ISM manufacturing, the Employment Cost Index, and initial jobless claims.
• Global macro indicators to watch, including China manufacturing PMI, Eurozone GDP and HICP, and Japan employment and consumption data.