AIORG-W006 Peer-reviewed research

Generative AI at Work

QJE qjae044 · NBER w31161

peer reviewed journal article

Recorded claims
2
coded from the inspected source
Evidence classes
1
scholarly-full-text
Recorded limitations
4
stated, never hidden

From the wave-0 evidence ledger

Recorded claims

Each claim states what the inspected source says, at the recorded location—bounded by its scope and graded by its confidence. Nothing here is a synthesis across works.

Across 5,172 customer-support agents, access to the generative assistant increased issues resolved per hour by about 15% on average, with substantially larger gains among novice and lower-skilled workers.

confidence: high within setting scholarly-full-text

Scope: Customer-support work in one organization; task productivity rather than firm profitability.

Performance convergence and evidence that newer workers learned from patterns associated with stronger workers support a knowledge-transfer mechanism; the study also reports improvements in customer sentiment and employee retention, but these outcomes remain setting-specific.

confidence: medium-high scholarly-full-text

Scope: Observed support-agent cohort and rollout period.

Boundaries

Limitations & independence

Recorded at coding time, carried with the work forever. A claim without its limits is not evidence.

Recorded limitations

  • One customer-support organization and task family
  • Rollout design is not a universal firm-level randomized trial
  • Product and organizational context may interact
  • Average effects conceal expertise heterogeneity

Source independence

Academic analysis of a staged workplace assistant rollout; focal-firm and product context but journal peer review.