Maxim Massenkoff
Anthropic economist whose research measures real AI impact on professional labour markets.
Last refreshed: 30 March 2026
Can Anthropic research on its own product be trusted to assess AI job displacement?
Latest on Maxim Massenkoff
- Who is Maxim Massenkoff?
- Maxim Massenkoff is an economist at Anthropic who co-authored a study on how AI tools such as Claude affect professional labour markets, introducing the concept of observed exposure to measure real workplace AI usage.Source: Anthropic
- What is observed exposure in AI labour research?
- Observed exposure measures the share of professional tasks a worker actually delegates to an AI tool, as opposed to theoretical capability-based frameworks. Massenkoff and McCrory found computer programmers face 75% task coverage under this measure.Source: Massenkoff & McCrory (Anthropic)
- Has AI caused unemployment in high-exposure jobs?
- Massenkoff and McCrory found no systematic unemployment increase in heavily AI-exposed occupations since late 2022, though they identified suggestive evidence of slowing hiring among workers under 25 in those roles.Source: Anthropic
- What did the American Enterprise Institute say about AI job displacement research?
- The AEI cited the Massenkoff-McCrory observed-exposure study to rebut Bernie Sanders' HELP Committee staff report, arguing the report ignores available data and that current AI tools function as skill equalisers raising performance at the bottom of the distribution.Source: American Enterprise Institute
- Can Anthropic research on AI job impacts be trusted?
- Critics note a conflict of interest when Anthropic researchers study the labour-market effects of Anthropic's own product. Massenkoff and McCrory acknowledge their methodology measures Claude usage specifically, though the paper has been cited across the political spectrum.Source: Anthropic / AEI
Background
Maxim Massenkoff is an economist and researcher at Anthropic, specialising in empirical labour economics. His work focuses on how large language models affect professional workflows and employment outcomes, particularly through the lens of actual usage data rather than theoretical exposure frameworks.
Massenkoff and Peter McCrory co-authored a study introducing "observed exposure", which measures real professional Claude usage against theoretical task-coverage models . Their findings show computer programmers face 75% task coverage; computer and mathematics occupations 35.8%; office and administrative roles 34.3%. The research characterises highly exposed workers as older, female, more educated and higher-paid, and finds no systematic unemployment increase in those occupations since late 2022, though hiring of younger workers is slowing.
That nuance sits at the heart of the AI employment debate: the study neither confirms catastrophic displacement nor dismisses structural risk. Its methodology has become a reference point in Congressional hearings and think-tank rebuttals , making Massenkoff an unusually prominent internal Anthropic voice in public labour-market policy.
