---
title: 'Accenture: GitHub Copilot-assisted daily engineering work'
slug: accenture-github-copilot-field-experiment
stable_id: bc88eb70918fc8a9
company: Accenture
function_code: software_engineering
pattern_codes:
  - creator_to_judge
evidence_strength: verified
publication_tier: showcase
freshness: current
reviewed_at: '2026-08-23'
updated_at: '2026-08-22'
source_quality_summary: 1 independent; publication outcomes are verified.
caveat_summary: >-
  Preliminary estimates were imprecise and statistically significant only under
  weighting for periods with stronger uptake differences.
collections:
  - human-still-decides
bundle_version: 1.0.0
bundle_fingerprint: sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93
canonical_url: https://brianletort.ai/transformations/accenture-github-copilot-field-experiment
---

# Accenture: GitHub Copilot-assisted daily engineering work

Developers can accept, edit, or reject inline suggestions while build, review, and merge controls remain.

Function: Software engineering. Patterns: Creator to judge. Evidence: verified.

Freshness: current. Reviewed: 2026-08-23. Updated: 2026-08-22.


Source quality: 1 independent; publication outcomes are verified.

## Before

1. **Developer** — Writes implementation and test code manually. (control: Tests and code review.)
2. **Reviewer** — Reviews submitted code and decides whether to merge. (control: Standard engineering governance.)

## After

1. **Developer with Copilot** — Receives inline code suggestions and accepts, edits, or rejects them. (control: Developer review, tests, scanning, and policy.)
2. **Reviewer** — Reviews the complete human-owned change. (control: Unchanged merge authority.)

## Decision rights

Developers decide whether suggestions enter a change; reviewers and repository controls decide merge.

## Exception path

Developers reject unsuitable output and code manually; failed tests or reviews block the change.

## Outcomes

- **Pull requests completed per week** (verified): Randomized Accenture developers without Copilot access. → Access was associated with 7.51%-8.69% more pull requests per week, depending on specification.. Preliminary estimates were imprecise and statistically significant only under weighting for periods with stronger uptake differences.

## Executive lesson

Real-world productivity evidence is more modest and noisier than lab claims; adoption and quality controls matter.

## Anti-pattern

Using pull-request count alone as proof of engineering value.

## Questions for leaders

- Did defect rates change?
- How much assigned access became actual use?
- Which repositories are excluded?

## Sources

- [The Productivity Effects of Generative AI: Evidence from a Field Experiment with GitHub Copilot](https://doi.org/10.21428/e4baedd9.3ad85f1c) — The Digital Economy Lab