AI’s Job Ledger Has Two Columns

Why It Matters

Understanding AI’s dual impact is crucial for policymakers and firms to design training, safety nets, and investment strategies that capture growth while mitigating displacement.

Key Takeaways

• AI exposure raised unemployment 0.3 pts above cycle baseline.
• Morgan Stanley sees 0.1% overall job loss from AI this year.
• Goldman Sachs estimates AI cut payroll growth by ~25,000 jobs monthly.
• LinkedIn data shows 640,000 AI‑related U.S. jobs created 2023‑2025.
• AI augmentation adds roughly 9,000 jobs per month, per Goldman.

Pulse Analysis

The conversation among economists about artificial intelligence and employment has moved from outright skepticism to cautious acknowledgement. Recent analyses from Morgan Stanley and Goldman Sachs reveal that AI‑exposed occupations are experiencing a modest but measurable uptick in unemployment—about 0.3 percentage points above what the business cycle alone would predict—translating into a 0.1% rise in the overall rate. Moreover, Goldman estimates a monthly reduction of roughly 25,000 jobs in payroll growth, echoing historical patterns where transformative technologies initially displace workers before productivity gains emerge.

Simultaneously, the creation side of AI’s ledger is gaining visibility. LinkedIn’s data indicate that between 2023 and 2025, roughly 640,000 AI‑related jobs were added in the United States, spanning roles such as model trainers, data annotators, and AI auditors—positions that often command premium wages. Augmentation effects further bolster employment, with Goldman estimating an addition of about 9,000 jobs per month as AI tools boost worker productivity across sectors. These emerging occupations, coupled with infrastructure demands like data‑center construction, suggest that AI is not merely a destroyer but also a catalyst for new labor markets.

For policymakers, the dual‑column ledger underscores the need for balanced strategies. Training programs that reskill displaced workers, mobility incentives, and robust safety nets can mitigate short‑term pain, while incentives for AI‑driven innovation can amplify job‑creation pathways. Ignoring either column risks a skewed response that either overstates the threat or underestimates the opportunity. As AI continues to mature, tracking both displacement and creation will be essential for shaping an economy that leverages higher productivity without leaving workers behind.