A major government statistical agency attributes net national employment decline directly to AI adoption before 2030.
Public debate centers on whether generative AI causes net job losses. This resolves if an official body like the US BLS reports net aggregate job losses driven by AI.
Across 6 frontier models answering independently, the panel has all but ruled it out: the median is 7%.
They broadly agree, 12 points between GPT at 18% and Claude at 6%.
Nothing is settled until 27 Sept 2031, when the outcome is checked against sources and every number above gets a permanent score.
What each one said
-
GPTOpenAI
It hinges on whether a top agency can credibly isolate AI’s net effect amid cyclical and policy-driven employment changes.
- BLS, Eurostat, or a peer would need to publish an explicit causal aggregate estimate by September 2031, not just occupation forecasts.
- Historically, statistical agencies report technology exposure and displacement but rarely attribute national net employment changes to one innovation.
- I would raise this sharply if BLS launches a recurring AI-adoption employer survey linked to employment outcomes by 2028.
18% -
DeepSeekDeepSeek
Agencies measure AI's labor effect indirectly and lag years; attributing net national decline to AI by 2030 is a high bar.
- BLS attributes job changes to business-cycle and industry factors, not technology; no precedent for an AI-caused net-decline finding.
- Any such claim needs multi-year CPS/JOLTS evidence plus a methodological shift, and BLS publishes with 1-2 year lags, leaving little runway.
- AI's measured employment effect so far is small and concentrated in specific occupations, not net national decline, so the base rate is very low.
9% -
GeminiGoogle
Statistical agencies measure aggregate labor counts, not causal macroeconomic drivers like specific technologies.
- Bodies like the BLS report net payroll changes, leaving causal attribution of technological displacement to academic literature.
- Macro net employment declines require recessions, during which agencies cite cyclical demand shock rather than AI adoption.
- Odds rise only if an agency establishes a novel causal tracking metric specifically mandated by federal statute before 2029.
7% -
GrokxAI
BLS/ONS-style agencies pinning a national net jobs drop on AI in an official release before 2030.
- No statistical agency has ever attributed a net national employment decline to a general-purpose technology; they cite cycles or demographics.
- AI effects remain sectoral through 2026; a true national net drop would require a recession they would not credit to AI without extraordinary evidence.
- A 2027–28 BLS CES or equivalent report naming AI as the primary driver of negative payrolls would flip this, but that bar is historically unmet.
7% -
QwenAlibaba
BLS methodology attributes job changes to industry/occupation shifts, not to a single technology cause, making direct AI attribution structurally unlikely.
- BLS and peer agencies (ONS, Eurostat) decompose employment by sector and occupation; they don't issue causal attributions to a single technology.
- net national employment decline itself is rare in developed economies outside recessions; even if one occurs, agencies blame macro cycles, not AI.
- a shift would require a methodological revolution or a political directive within 5 years—no such proposal is on BLS's published research agenda.
7% -
ClaudeAnthropic
Requires BLS/ONS/Eurostat to formally attribute net aggregate job losses to AI, which agencies rarely do causally.
- Statistical agencies report descriptive employment/occupation data but almost never issue causal attribution of net job losses to a single technology.
- Even during clear automation waves (e.g. manufacturing robotics), agencies avoided direct causal claims, favoring correlational occupational breakdowns instead.
- A sharp AI-driven recession combined with unusual political pressure for agencies to name AI explicitly would be the main path to resolution.
6%
6 of 6 models answered · 12 points between the highest and lowest. None was shown the market price.
Question sourced from a news sweep on 28 Sept 2026. Forecast by google/gemini-3.8-flash, anthropic/claude-sonnet-5, openai/gpt-5.6-terra, x-ai/grok-4.6, deepseek/deepseek-v4.1-flash, qwen/qwen3.8-max-0902 via OpenRouter.