Harvard researchers find AI coding agents shifted the bottleneck to code review
A Harvard job-market working paper by Fiona Chen and James Stratton, covering 718 firms, finds that engineers at firms adopting AI coding agents added about 30 percent more code, while finished work did not measurably rise.
Source-based brief · Editor review not recorded · See checks and limits ↓
One primary source; other links are context, not independent corroboration.
- Primary sourcecomplexdiscovery.com
- Discovery linkHacker News
- Linked in sourceArtificial Intelligence in the Firm: Bottlenecks in Software Production
- Linked in sourceJames Stratton | Department of Economics
- Linked in sourceWriting Code vs. Shipping Code: Productivity Effects Across Generations of AI Coding Tools
- Linked in sourceRule 11. Signing Pleadings, Motions, and Other Papers; Representations to the Court; Sanctions
- Linked in sourceIntroducing the Motion to Dismiss Workflow, Co-Developed With LexisNexisⓇ
- Linked in sourceCooley Collaborates With Google Cloud to Develop AI Agent for Complex Litigation
- Linked in sourceJellyfish Reveals AI’s Real Impact on Engineering Teams with Industry’s Largest Ongoing Benchmark Study
- Linked in sourceNew Research from Harvard & Jellyfish: AI Is Making Developers Faster. So Where’s the Business Impact?
- Linked in sourceDoes AI Assistance Mean Engineers Are Doing Less Coding Work? Maybe Not.
- Linked in sourceAI Coding Agents Productivity Gains Offset by Human Review
- Linked in sourceCooley and Google Cloud are building a court-filing redaction agent without a published error rate
- Linked in sourceMonsterCloud indictment exposes the ransom payment a victim’s own file may not record
A community post points to the original link. Discussion and upvotes do not establish accuracy.
Open full article view →- Community discussion+5
- Fetched source page+30
- No editor link review+0
- Community link+0
This measures the stored source trail and editorial steps. It does not certify accuracy, peer review, or independent corroboration. Popularity adds no points.
View full scoring →Read the scoring rubric →What the source says
A Harvard job-market working paper by Fiona Chen and James Stratton, covering 718 firms, finds that engineers at firms adopting AI coding agents added about 30 percent more code, while finished work did not measurably rise. Time from pull request submission to merge rose 49 percent, and code review became the bottleneck. The paper has not completed peer review, and the study examines software engineering rather than legal work.
Claim explorer
Attribution, source and conditionsSource-stated points from the stored brief. Claim-level assessments have not been recorded.
The paper is a job-market working paper without completed peer review, and the study covers software engineering, not legal work. Source claims not independently checked.
This connects a source, an attributed statement and its qualifications. It does not establish that the claim is correct.
Questions worth asking
Open questions for further reportingWhat remains unconfirmed?
The paper is a job-market working paper without completed peer review, and the study covers software engineering, not legal work. Source claims not independently checked.
What material was available for this brief?
The recorded brief basis is fetched source page. These points are AINewsIQ paraphrases, not quotations or independent confirmations.
Lines of code added rose about 30 percent per worker-month after agent adoption at adopting firms.
Source: complexdiscovery.com ↗Stored source-based brief · key point 1 · AINewsIQ paraphraseFinished work, measured as Jira issues resolved, showed no statistically detectable increase.
Source: complexdiscovery.com ↗Stored source-based brief · key point 2 · AINewsIQ paraphraseTime from pull request submission to merge rose 3.45 days, a 49 percent increase over baseline.
Source: complexdiscovery.com ↗Stored source-based brief · key point 3 · AINewsIQ paraphraseAI review tools generated only 23.3 percent of review comments by March 2026, so the bottleneck persisted.
Source: complexdiscovery.com ↗Stored source-based brief · key point 4 · AINewsIQ paraphraseLimits from the stored brief
The paper is a job-market working paper without completed peer review, and the study covers software engineering, not legal work. Source claims not independently checked.
This is a source-based abstract, not independent validation of the source's claims.
Evidence & sources
These links show where the signal came from and what the source page linked to. A linked page is context, not independent corroboration.
A community post points to the original link. Discussion and upvotes do not establish accuracy.
Timeline & updates
- Signal dated by source or discovery feed
This timestamp places the signal in the radar. It may be the discovery post date rather than the primary page's first publication date.
- Shared brief stored
AINewsIQ prepared one abstract from fetched source page.
Brief history · 1 retained version
These are changes to AINewsIQ’s own publication. Earlier history is unavailable unless recorded here. Complete retained versions are available to the editor.
- Version 1 · published
Source-based brief stored or updated. This does not establish that the publisher changed its page.
Automated record
Source observations · 1 recent captures
A changed capture means the extracted text differed. It can reflect page furniture or extraction changes; it does not establish a substantive correction. Source captures remain private.
Retained capture; no earlier comparison available.
New engagement snapshots do not mean the source itself changed. Source observations and AINewsIQ brief revisions are separate records.
Why this may matter
The findings suggest that faster drafting can shift the constraint to human review, so buyers of AI tools may need to measure review-stage time and revision rates, not only output volume.
Limits and open questions
The paper is a job-market working paper without completed peer review, and the study covers software engineering, not legal work. Source claims not independently checked.
Organization-specific financial, security, and operational effects require the organization's own context and review.

Join the discussion
Have a different perspective or a question about the evidence? Discussion is free with a verified email account. Every comment is reviewed before publication.
No published comments yet.
New commenters: 2 submissions per 24 hours, at least 5 minutes apart, until 3 are approved. Discussion guidelines and moderation →