Claim analyzed

Tech

“Generative artificial intelligence assistance can increase worker productivity, with larger productivity gains for less-experienced workers.”

Submitted by Witty Lark 19cb

True
9/10

Controlled experiments and workplace field studies show that generative AI assistance can improve productivity, often with the largest gains among novice, junior, or lower-performing workers. Effects vary by task and implementation, and some studies find limited or negative effects for experienced workers, but those qualifications do not contradict the claim's carefully bounded wording.

Caveats

  • Productivity effects depend on the task, tool, workflow, and performance metric used.
  • AI can reduce quality or provide little benefit when tasks exceed its capabilities, especially without effective oversight.
  • Several cited pages summarize or reproduce the same underlying studies, so the source list overstates the number of independent findings.

Sources

Sources used in the analysis

#1
bis.org 2024-09-01 | Generative AI and labour productivity: a field experiment on coding

Our findings indicate that the use of gen AI increased code output by more than 50%. However, productivity gains are statistically significant only among entry-level or junior staff, while the impact on more senior employees is less pronounced.

#2
nber.org 2023-04-01 | April 2023, revised November 2023

Access to the tool increases productivity, as measured by issues resolved per hour, by 14% on average, including a 34% improvement for novice and low-skilled workers but with minimal impact on experienced and highly skilled workers.

#3
nber.org 2023-06-01 | Measuring the Productivity Impact of Generative AI | NBER

Customer support agents using an AI tool to guide their conversations saw a nearly 14 percent increase in productivity, with 35 percent improvements for the lowest skilled and least experienced workers, and zero or small negative effects on the most experienced/most able workers, Erik Brynjolfsson, Danielle Li, and Lindsey R. Raymond report in Generative AI at Work(NBER Working Paper 31161).

#4
nber.org 2024-04-24 | The Economics of Generative AI | NBER

On average, worker productivity increased by 14 percent. These gains were concentrated among the lowest quintile of workers, whether measured by experience or prior productivity, where there were productivity gains of up to 35 percent.

#5
science.org Experimental evidence on the productivity effects of generative artificial intelligence

Participants assigned to use ChatGPT were more productive, efficient, and enjoyed the tasks more. … Participants with weaker skills benefited the most from ChatGPT, which carries policy implications for efforts to reduce productivity inequality through AI.

#6
gsb.stanford.edu 2025-05-01 | Generative AI at Work: Journal Article | Stanford Graduate School of Business

Access to AI assistance increases worker productivity, as measured by issues resolved per hour, by 15% on average, with substantial heterogeneity across workers. … Less experienced and lower-skilled workers improve both the speed and quality of their output, while the most experienced and highest-skilled workers see small gains in speed and small declines in quality.

#7
nber.org 2023-04-24 | Generative AI at Work

Customer support agents using an AI tool to guide their conversations saw a nearly 14 percent increase in productivity, with 35...

#8
digitaleconomy.stanford.edu 2023-04-15 | Generative AI at Work - Stanford Digital Economy Lab

Access to AI assistance increases worker productivity, as measured by issues resolved per hour, by 15% on average, with substantial heterogeneity across workers. Less experienced and lower-skilled workers improve both the speed and quality of their output while the most experienced and highest-skilled workers see small gains in speed and small declines in quality.

#9
siepr.stanford.edu 2023-04-01 | Generative AI at Work | Stanford Institute for Economic Policy Research (SIEPR)

Access to the tool increases productivity, as measured by issues resolved per hour, by 14% on average, including a 34% improvement for novice and low-skilled workers but with minimal impact on experienced and highly skilled workers. … Our results suggest that access to generative AI can increase productivity, with large heterogeneity in effects across workers.

#10
microsoft.com 2025-08-19 | The Effects of Generative AI on High-Skilled Work: Evidence from Three Field Experiments with Software Developers - Microsoft Research

Though each experiment is noisy, when data is combined across three experiments and 4,867 developers, our analysis reveals a 26.08% increase (SE: 10.3%) in completed tasks among developers using the AI tool. Notably, less experienced developers had higher adoption rates and greater productivity gains.

#11
economics.mit.edu 2023-03-02 | Experimental Evidence on the Productivity Effects of

Our results show that ChatGPT substantially raises average productivity: time taken decreases by 0.8 SDs and output quality rises by 0.4 SDs. … Inequality between workers decreases, as ChatGPT compresses the productivity distribution by benefiting low-ability workers more.

#12
microsoft.com 2024-09-06 | Generative AI at Work - Microsoft Research

Access to the tool increases productivity, as measured by issues resolved per hour, by 14 percent on average, with the greatest impact on novice and low-skilled workers, and minimal impact on experienced and highly skilled workers.

#13
ideas.repec.org 2025-01-01 | Generative AI at Work

Access to AI assistance increases worker productivity, as measured by issues resolved per hour, by 15% on average, with substantial heterogeneity across workers. … Less experienced and lower-skilled workers improve both the speed and quality of their output, while the most experienced and highest-skilled workers see small gains in speed and small declines in quality.

#14
papers.ssrn.com 2024-11-06 | Generative AI in Action: Field Experimental Evidence on Worker Performance in E-Commerce Customer Service Operations

Results show that the gen AI assistant significantly enhanced both service speed and service quality. … Analysis by agents' pretreatment performance reveals that low performers experienced greater improvements in speed and quality, narrowing the performance gap, while high performers saw a decline in service quality, likely because gen AI suggestions fell below their expertise.

#15
hbs.edu 2023-09-01 | Navigating the Jagged Technological Frontier: Field Experimental Evidence of the Effects of Artificial Intelligence on Knowledge Worker Productivity and Quality - Working Paper - Faculty & Research - Harvard Business School

For each one of a set of 18 realistic knowledge tasks within the frontier of AI capabilities ranging from creative to analytical tasks, subjects using AI outperformed those not using AI, completing 12.2% more tasks and completing them 25.1% more quickly on average while also delivering solutions of significantly improved quality.

#16
arxiv.org 2023-04-23 | Generative AI at Work

Access to AI assistance increases worker productivity, as measured by issues resolved per hour, by 15% on average, with substantial heterogeneity across workers. … Less experienced and lower-skilled workers improve both the speed and quality of their output while the most experienced and highest-skilled workers see small gains in speed and small declines in quality.

#17
academic.oup.com Generative AI at Work* | The Quarterly Journal of Economics
#18
oecd.org OECD Employment Outlook 2025 | OECD
#19
oecd.org Generative AI boosts the performance of beginners
#20
oecd.org OECD Employment Outlook 2025: Generative AI boosts the performance of beginners
#21
danielle.li 2025-02-04 | GENERATIVE AI AT WORK*

Access to AI assistance increases worker productivity, as measured by issues resolved per hour, by 15% on average, with substantial heterogeneity across workers. … Less experienced and lower-skilled workers im prove both the speed and quality of their output, while the most experienced and highest-skilled workers see small gains in speed and small declines in quality.

#22
hai.stanford.edu The First Real-World Study of Generative AI at Work Finds a 14% Productivity Boost | Stanford HAI
#23
hai.stanford.edu AI Assistance Improves Customer Service Worker Productivity and Sentiment
#24
hai.stanford.edu The First Real-World Study of Generative AI at Work Finds Increased Productivity and Happiness
#25
iwer.mit.edu 2022-03-02 | How does access to a generative AI tool affect work in a call center?

IWER Generative AI and Worker Productivity … Ideas Made to Matter How generative AI can boost highly skilled workers’ productivity

Full Analysis

Debate

Two AI advocates debated this claim using the research gathered.

Argument for

P
Proponent Argues TRUE

Verified field evidence from the BIS finds generative AI raised code output by more than 50%, with statistically significant gains concentrated among entry-level and junior staff (Source 1, bis.org). Independently, the NBER customer-service study reports a 14% average productivity increase but a 34% improvement for novice/low-skilled workers and minimal effects for experienced workers, directly establishing the claim's larger gains for less-experienced workers (Source 2, nber.org).

O
Opponent Rebuttal

The Proponent's selective emphasis on Source 1's overall code-output rise and Source 2's 14% average ignores those same sources' explicit findings that gains are statistically insignificant or minimal for senior and experienced workers, rendering any general claim of increased worker productivity false. By treating task-specific, heterogeneous results as universal support, the Proponent commits a hasty generalization that the research brief's consistent evidence of quality declines among high-skilled staff (Sources 6 and 8) directly refutes.

Argument against

O
Opponent Argues FALSE

The claim is false because multiple high-quality studies reveal that generative AI produces no meaningful or even negative productivity effects for experienced workers, with Source 6 (gsb.stanford.edu) and Source 8 (digitaleconomy.stanford.edu) documenting small quality declines among the highest-skilled, while Source 1 (bis.org) finds gains statistically insignificant outside juniors. These task-specific field experiments further undermine any general assertion of productivity increases, as Source 14 (papers.ssrn.com) shows high performers suffering quality drops that offset speed gains and expose the claim as overstated and contextually misleading.

P
Proponent Rebuttal

The Opponent commits a straw man by treating the claim that generative AI can increase productivity, with larger gains for less-experienced workers, as though it asserted meaningful gains for experienced workers; Sources 1, 2, 6, and 8 instead directly support the stated heterogeneity. The Opponent's assertion that quality declines “offset” speed gains is unsupported by Source 14, whereas Source 1 reports more than 50% higher code output concentrated among juniors and Source 2 finds a 14% average gain and a 34% novice/low-skilled gain.


Panel Review

3 specialized AI experts evaluated the evidence and arguments.

Reviewer 1 — The Logic Examiner

Focus: Inferential Soundness & Fallacies
True
9/10

The claim has two components: (1) generative AI can increase productivity, and (2) the gains are larger for less-experienced workers — both are directly and repeatedly supported by convergent evidence across independent field experiments (Sources 1, 2, 5, 6, 8, 10, 11, 12, 13, 14), each showing average productivity gains alongside disproportionately larger improvements for novice/low-skilled workers and minimal or negative effects for experts. The Opponent's rebuttal commits a hasty generalization/straw man by treating 'minimal impact on experts' as refuting the claim, when the claim never asserts uniform gains for all workers — it explicitly asserts differential gains favoring less-experienced workers, which is precisely what the evidence shows; thus the logical chain from evidence to claim is direct and sound, and the claim is true.

Logical fallacies

The Opponent commits a straw man by interpreting the claim as asserting universal productivity gains for all workers, when it explicitly specifies larger gains for less-experienced workers.The Opponent's rebuttal also reflects a hasty generalization by treating isolated quality-decline findings among high performers as sufficient to negate the well-supported average productivity increase documented across many independent studies.
Confidence: 9/10

Reviewer 2 — The Source Auditor

Focus: Source Reliability & Independence
True
10/10

Multiple highly reliable sources, including NBER (Sources 2, 3, 4, 7), BIS (Source 1), Science (Source 5), and Stanford (Sources 6, 8, 9), consistently confirm that generative AI increases average worker productivity while disproportionately benefiting less-experienced or lower-skilled workers. The evidence pool is robust, recent, and independent, clearly validating both parts of the claim.

Confidence: 9/10

Reviewer 3 — The Precision Analyst

Focus: Claim Precision & Quantitative Accuracy
True
10/10

The causal wording is supported by field and experimental evidence: Sources 1 and 2 report productivity increases from generative-AI access or use, and both find gains concentrated among junior, novice, or lower-skilled workers; Sources 5, 6, and 8 independently report the same direction of heterogeneity. Because the claim says AI assistance "can" increase productivity rather than asserting universal gains across every worker or task, the minimal or negative effects for experienced workers do not contradict it.

Confidence: 9/10

Panel summary

See the full panel summary

Create a free account to read the complete analysis.

Sign up free
The claim is
True
9/10
Confidence: 9/10 Spread: 1 pts

Your annotation will be visible after submission.

Embed this verification

Every embed carries schema.org ClaimReview microdata — recognized by Google and AI crawlers.

True · Lenz Score 9/10 Lenz
“Generative artificial intelligence assistance can increase worker productivity, with larger productivity gains for less-experienced workers.”
25 sources · 3-panel audit · Verified Aug 2026
See full report on Lenz →