Trends›LLMs lift productivity 20% by 2030
LLMs lift productivity 20% by 2030
Large language models raise OECD labour productivity by 20% by 2030 versus a 2022 baseline, showing up in faster macro growth, lower prices and higher living standards.
17 confidence, computed 23 Oct 2026
- Social
Evidencethe schedule the figure draws
The evidence, newest first
- Jun 2026
against · primary OECD labour productivity growth remained subdued at 1.2% weighted average (0.4% median) in 2024, with multifactor productivity contributing negatively in 14 of 21 countries; AI impact not yet evident in productivity statistics. OECD Compendium of Productivity Indicators 2026 - Labour Productivity Trends - Jun 2026
against · primary Nonfarm business sector labor productivity increased 0.3 percent in Q1 2026 (seasonally adjusted annualized rate) and 2.8 percent year-over-year, below long-term 2.1% trend. Productivity and Costs, First Quarter 2026, Revised - May 2026
against · observable Overall AI usage hovered between 17% and 20% from December 2025 to May 2026, with 37% of firms with 250+ employees using AI. Large Firms With at Least 20 Employees Biggest AI Users - May 2026
neutral · observable Total factor productivity increased only 1.1 percent from 2019 to 2024 despite rapid AI investment growth; rapid software investment growth from 2021-2024 may indicate AI usage but productivity gains remain modest. AI and the rise of software investment - May 2026
for · observable Nonfarm business productivity rose 0.3% in Q1 2026 (annualized) and 2.8% YoY; current-cycle annualized rate 2.1%, above the prior cycle's 1.5%. BLS Productivity and Costs Q1 2026 - May 2026
against · primary Total factor productivity fell from 1.5% in 2024 to 0.8% in 2025, with efficiency gains not contributing to labor productivity growth in 2025 as much as in 2024; AI contributing through capital spending, not yet through efficiency. Q1 2026 Productivity and Costs Release: Productive, for Now - May 2026
against · observable Nonfarm business productivity grew 0.8% annualized in Q1 2026, a sharp deceleration from 2025 pace, with more output growth coming from labor input rather than efficiency. Why Your 2026 Raise Just Got Smaller: US Productivity Numbers - Apr 2026
against · primary During Nov 2025-Jan 2026, 18% of firms used AI in a business function, rising to 32% on an employment-weighted basis. The Microstructure of AI Diffusion: Evidence from Firms, Business Functions, and Worker Tasks - Apr 2026
neutral · observable 18% of firms had adopted AI by end-2025 and 41% of the workforce used generative AI at work as of Nov 2025, but no direct productivity measurement — the payoff treated as an open question. Fed FEDS Note — Monitoring AI adoption - Mar 2026
neutral · observable U.S. Total Factor Productivity 2025 - Mar 2026
neutral · claim AI-related capex accounted for roughly 75% of US economic growth in Q1 2026 — a demand/investment effect, not measured productivity, and a reversal risk if the build-out slows. AI capex ~75% of US growth Q1 2026 (analyst decomposition) - Feb 2026
for · observable Industry analysis finds positive AI adoption-growth link with moderate fit (R²=0.19), and nonfarm business labor productivity accelerated in 2025:Q3 relative to 2025:Q2, with pickup above pre-pandemic trend since late 2022. A New U.S. Productivity Chapter? What Industry Data Say About AI - Sep 2025
against · primary Projects AI adds only ~0.09pp to TFP in 2027 and ~0.18pp by 2030, cumulative GDP/productivity just ~1.5% higher by 2035 — far short of +20% by 2030. Penn Wharton Budget Model — AI productivity projection - Aug 2025
against · observable MIT Project NANDA found ~95% of enterprise generative-AI pilots delivered no measurable P&L return despite $30-40bn invested — a diffusion bottleneck between adoption and realized productivity. MIT: 95% of enterprise gen-AI pilots failing (Fortune) - Jun 2025
against · primary OECD-wide labour productivity grew only ~0.6% in 2023 and ~0.4% in 2024; MFP stagnated or fell in most countries and 'AI's impact is not yet evident in the productivity statistics.' OECD Compendium of Productivity Indicators 2025 - Jan 2025
against · primary Estimates total AI-driven TFP gains of no more than 0.53-0.66% over ten years (~0.064%/yr), an order of magnitude below a 20%-by-2030 path. Acemoglu, The Simple Macroeconomics of AI (NBER w32487)
What would move itthe measures watched
- watchBLS nonfarm business labor productivity (output per hour), quarterly + YoY. Sustained acceleration above the ~2.1% post-1947 average toward 3%+ would be the first hard macro confirmation; a stall toward 1-1.5% is disconfirming
- watchOECD total-economy labour productivity (annual, all-OECD + euro area). The cross-country aggregate is the actual scoring surface; OECD itself flags AI is 'not yet evident'
- watchCensus BTOS firm AI-use rate (overall + employment-weighted). Adoption is the leading input; rising adoption with flat productivity = diffusion lag
- watchAcademic/official TFP-contribution estimates (Acemoglu, Penn Wharton, Fed). Compare central-case annual TFP boost (~0.06-0.2pp) against the ~2.3%/yr a 20%/8yr path needs
- watchAI capex share of US GDP growth. Capex carrying GDP is NOT the productivity channel; it can mask weak productivity and reverse if the build-out slows