AI's Impact on Employment

AI's impact on employment is mixed

Studies show negative effects on low-skill jobs but positive effects on high-skill positions. Among 40 studies, 20 found positive impacts, 15 negative, and 11 neutral.

Abstract

Artificial intelligence exhibits a heterogeneous impact on employment. Several studies report declines in low-skill and routine roles, such as:

• 21% drop in job posts for writing and coding tasks.
• 2% decline in freelance jobs.
• 5.2% reduction in earnings.

Conversely, other work documents gains among high-skill workers:

• Estimates include a 2.6–4.3% increase in high-skill job shares.
• Some papers report no aggregate change, as sectoral shifts offset one another, leading to longer-term job transformation after short-term displacement.

Overall, among 40 studies, the direction of AI’s impact is:

• Positive in 20 studies
• Negative in 15 studies
• Neutral in 11 studies
  Effects vary by worker type, industry, and region, with declines often offset by new opportunities in services and technology-driven sectors.

Methods

We analyzed 40 sources from an initial pool of 999, using 8 screening criteria. Each paper was evaluated for:

• AI-Employment Relationship
• Quantitative Employment Measures
• Empirical Methodology
• AI Technology Focus
• Empirical Data Presence
• AI-Specific Content
• Comparison or Baseline
• Publication Type

Papers included for extraction: 40

Paper search

Using the research question “What is the estimated impact of AI on employment?”, we searched across over 126 million academic papers. 999 papers most relevant to the query were retrieved.

Screening

Inclusion criteria:

• Examining AI technologies and employment outcomes.
• Measuring quantitative employment impacts.
• Using empirical study methodologies.
• Examining specific AI technologies.
• Containing empirical data.
• Focusing on AI technologies directly.
• Including comparison groups or baseline measurements.
• Being a full research article.

Data extraction

We extracted data columns based on the following:

• Study Design:
  Identify the primary research design used in the study (e.g., panel data analysis, cross-country econometric analysis).
• Data Sources and Sample Characteristics:
  Extract details such as geographic coverage, time period, and specific data sources.
• AI Exposure Measurement Method:
  Describe how the study measured AI’s impact on employment including metrics and classification methods.
• Key Employment Outcomes:
  List primary employment-related outcomes examined (employment growth, job numbers, wage impacts).
• Primary Findings on AI-Employment Relationship:
  Summarize the study’s conclusions about AI’s impact on employment, noting overall effects and mechanisms.

Results

Characteristics of Included Studies

Summary of Methodologies

• Panel data methods: 25 out of 40 studies.
• Cross-sectional/survey-based methods: 3 studies.
• Time series analysis: 3 studies.

Summary of Samples/Data

Geographic Coverage: US, China, OECD countries, etc.

Summary of Time Periods

Coverage spans from the 1980s to early 2020s, most studies focused on 2010-2025.

Effects of AI on Employment

Overall Employment Effects

Key Patterns

• Both positive and negative effects are reported, frequently depending on skill level, sector or other subgroups.
• Statistical significance was reported in less than half of the studies.

Differential Effects by Worker Type

• Low-skilled and routine workers negatively affected.
• High-skilled and STEM workers benefited from AI adoption.

Sectoral and Firm-Level Variations

• Service sector: generally positive or neutral effects.
• Industrial/manufacturing sector: more likely to experience job losses.

Temporal Dynamics and Adjustment Processes

• Short-term displacement followed by potential job creation and skill upgrading.

Synthesis and Limitations

• Evidence base: extensive and methodologically diverse.
• Heterogeneity complicates direct comparison and generalization.
• Context-dependent impact of AI on employment.