AI is boosting US output rather than significantly cutting jobs

Overview

This report is Morgan Stanley's US economics weekly, with the core theme being the impact of AI on US productivity, output, and the labor market. Using industry-level output-per-employee metrics to measure labor productivity and grouping industries by AI exposure, the report finds that since 2025, highly AI-exposed industries have shown more prominent productivity growth and acceleration. The report also updates US financial conditions, tariffs, 1Q GDP tracking, PCE inflation, upcoming macro data, and the US economic outlook.

Core views

The core view is that AI currently appears to be boosting output and income potential rather than causing significant job cuts. Productivity gains in highly AI-exposed industries have been accompanied by faster output growth, while employment growth has not weakened meaningfully on a relative basis. Morgan Stanley believes AI-related investment has already supported productivity in AI-related industries, while AI applications are also helping other industries improve production processes. However, as AI adoption rises, the pattern of labor disruption may change in the future, so output, employment, and AI exposure should continue to be tracked by industry.

Analysis framework

The report first constructs an industry output-per-employee metric using BEA industry output data and BLS employment figures, then uses the AI exposure standard from Felten et al. (2021) to divide industries into high, medium, and low AI exposure groups, comparing productivity growth and acceleration across groups in 2025. In the macro section, it also uses the FRB/US model to convert asset price changes into equivalent changes in the federal funds rate, and mechanically aggregates monthly activity data to form a 1Q GDP tracking estimate.

Methodology notes

• Productivity AnalysisIndustry output per employee

  Proxy metric for labor productivity

  BEA industry output data is divided by BLS employment to construct industry-level output per employee, used to observe the relationship between AI exposure and productivity growth.

• AI Exposure GroupingFelten et al. (2021) AI exposure criterion

  Grouping industries by high, medium, and low AI exposure

  The report classifies industries into the top 25%, middle 50%, and bottom 25% by AI exposure, and compares changes in output, employment, and productivity across groups.

• Financial ConditionsFRB/US-based financial conditions model

  Policy-rate-equivalent impact of asset price changes

  The model incorporates the 10-year US Treasury yield, S&P 500 returns, BBB credit spreads, US dollar valuation, and oil prices, and converts them into equivalent federal funds rate changes based on their growth elasticities.

• GDP TrackingMorgan Stanley 1Q 2026 GDP tracking

  Mechanical aggregation of monthly activity data

  The GDP tracking estimate differs from the official GDP forecast, relying mainly on monthly activity data that can directly enter BEA accounting to mechanically aggregate 1Q growth.

Asset mapping & comparison

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

• US Macroeconomy

  Core research subject

  Strengths

  AI-related productivity improvement, upward revision to 1Q GDP tracking, and continued support from consumption and nonresidential fixed investment.

  Weaknesses

  Tariffs, oil prices, and inflation pressure real spending, while delays in some government data increase uncertainty around GDP estimates.

  Comparison

  Morgan Stanley's 1Q GDP tracking is 2.4%, higher than the Atlanta Fed's tracking estimate of about 1.2%-1.3%.

  Risks

  Sticky inflation, tariff uncertainty, renewed tightening in financial conditions, and geopolitical shocks.

• US Labor Market

  Object for observing AI impact

  Strengths

  Employment growth trends are similar across industries with different AI exposure, with no broad evidence yet of widespread labor substitution.

  Weaknesses

  The current pattern may change as AI adoption increases.

  Comparison

  Productivity improvement in highly AI-exposed industries mainly comes from output growth rather than employment cuts.

  Risks

  Future diffusion of automation could widen employment divergence across industries.

• US Rates and Financial Conditions

  Policy transmission and market shock channel

  Strengths

  Financial conditions eased by 42bp after the April 7 ceasefire announcement.

  Weaknesses

  After February 28, financial conditions were at one point equivalent to about a 30bp rate hike.

  Comparison

  The main drivers were higher 10-year US Treasury yields and a stronger dollar, with oil prices as a secondary factor.

  Risks

  Higher Treasury yields, a stronger dollar, rising oil prices, and wider credit spreads could restrain growth.

• AI-related industries

  Main source of productivity improvement contribution

  Strengths

  Highly AI-exposed industries contributed more to productivity growth in 2025, while AI-related investment supported capital deepening.

  Weaknesses

  The report does not provide investment ratings for any single industry or stock.

  Comparison

  Output-per-employee growth in highly AI-exposed industries is faster than in medium- and low-AI-exposure industries.

  Risks

  Returns on AI investment, adoption speed, and labor substitution pathways remain uncertain.

Key data

• Contribution of highly AI-exposed industries to productivity growth1.7 percentage pointsOver the four quarters through 4Q 2025, highly AI-exposed industries contributed about 1.7 percentage points of the 2.4 percentage points in productivity growth.
• Contribution of highly AI-exposed industries in 20240.7 percentage pointsThe report notes that the contribution of highly AI-exposed industries to productivity growth accelerated in 2025 versus 2024.
• Impact of AI adoption on unemployment rateAbout 0.1 percentage point upward pressurePrevious research suggests that AI adoption has so far caused only limited disruption to the labor market.
• Impact of financial conditions after the Middle East conflictRoughly equivalent to a 30bp increase in the federal funds rateSince hostilities began on February 28, the tightening in financial conditions has mainly been driven by a rise in the 10-year US Treasury yield and US dollar appreciation, with higher oil prices as a secondary factor.
• Change in financial conditions after the April 7 ceasefire announcement42bp easingThe report says financial conditions eased noticeably after the ceasefire announcement.
• Effective US tariff rate in February 2026About 8.5%The report estimates the effective tariff rate in the February data at about 8.5%.
• Baseline tariff estimateClose to 11%Estimate after replacing IEEPA tariffs with a 15% Section 122 tariff and accounting for the impact of import composition.
• Core tariff estimateAbout 13%-14%Core tariffs exclude fuel, gold, and AI-related imports.
• 1Q GDP tracking2.4%Strong retail sales and upward revisions to earlier data raised 1Q GDP tracking by 0.2 percentage points to 2.4%.
• 1Q core PCE price4.1% q/q annualizedThe report expects 1Q core PCE prices to rise at an annualized quarterly rate of 4.1%, with headline PCE at 4.3%.

Impact & implications

If the report's judgment is correct, the near-term macroeconomic impact of AI is more positive on the supply side: faster productivity and output growth can support labor income or corporate income without immediately showing up as large-scale job substitution. For policymakers and markets, this mix helps improve the potential growth narrative, but constraints from tariffs, oil prices, financial conditions, inflation, and geopolitics on consumption, business investment, and the Fed's policy path still need attention.

Risks

• Middle East conflict and rising oil prices could tighten financial conditions again and suppress real consumption.
• Tariff policy remains uncertain: after the 150-day limit of Section 122, congressional extension will be needed, while Section 232 and 301 investigations will also affect subsequent tariff rates.
• 1Q GDP estimates face higher-than-normal uncertainty due to delays in construction spending and housing starts data.
• The impact of AI adoption on the labor market is currently small, but as adoption rises, industry employment disruption could increase.
• Core PCE and headline PCE are both expected to be elevated on an annualized basis in 1Q, which may limit room for rapid Fed easing.

What to watch

• Whether output, employment, and output per employee in highly AI-exposed industries continue to lead.
• The actual performance of the US 1Q GDP initial reading, consumption, nonresidential fixed investment, and the rebound in government output.
• March personal income, personal spending, PCE, and core PCE data.
• Whether the FOMC keeps rates unchanged, and how the statement and press conference describe inflation, growth, and uncertainty.
• The effective US tariff rate, follow-up arrangements for Section 122 tariffs, and the progress of Section 232 and 301 investigations.
• The joint movements of the 10-year US Treasury yield, the US dollar, oil prices, credit spreads, and US equities within the financial conditions index.