evidoria

← Back to browse

Good practice Imported

San Francisco's Citywide Microsoft 365 Copilot Rollout — From a 2,000-Employee Pilot to 30,000 City Workers

United States of America · San Francisco · See the United States of America profile · See the San Francisco profile

Evidence: Descriptive / self-reported Top 32% 67/100 · Ask Evidence Copilot about this practice

San Francisco scaled Microsoft 365 Copilot from a six-month, 2,000-employee pilot with self-reported productivity gains to nearly 30,000 city workers, logging over 1 million interactions by April 2026 - without independent verification of the productivity claims.

2,000+ staff
Pilot participants (Jan-Jun 2025)
up to 5 hours/week
Self-reported productivity gain
~30,000 employees
Citywide Copilot access (mid-2026)
40+
Departments covered
1,000,000+
Copilot interactions logged (Jul 2025-Apr 2026)
San Francisco's Citywide Microsoft 365 Copilot Rollout — From a 2,000-Employee Pilot to 30,000 City Workers

Details

Maturity
Scaling
Promoter
City and County of San Francisco — Office of the Mayor and Department of Technology, with Microsoft and InnovateUS
Period
Pilot: ~January-June 2025 (2,000+ staff); citywide rollout: July 2025 - ongoing (30,000+ employees by mid-2026)
Keywords
generative AI, workforce productivity, local government, digital services

Context

San Francisco sought to bring generative AI to its workforce under Mayor Daniel Lurie's technology agenda, starting with a six-month pilot involving more than 2,000 staff across departments including public health and social services.

Objectives

To test Microsoft 365 Copilot Chat (built on OpenAI's GPT-4o) for everyday tasks such as email drafting, project planning, research and document summarisation, then decide whether to scale citywide.

Activities

Following the pilot, the city rolled out Copilot from July 2025, pairing it with a five-week training programme delivered with InnovateUS and updated Generative AI Guidelines (8 July 2026) under the pre-existing 22J AI Transparency Ordinance. The rollout used San Francisco's existing Microsoft enterprise licensing.

Results

Pilot participants self-reported productivity gains of up to five hours per week. By mid-2026, access had scaled roughly fifteen-fold to nearly 30,000 employees across more than 40 departments, with over one million Copilot interactions logged between July 2025 and April 2026.

Conclusions

The scale of adoption and the transparency-ordinance safeguards are well documented, but the five-hours-per-week productivity figure comes from self-reported pilot surveys rather than an independent time-and-motion study, and no citywide before/after service-delivery metrics have been published - the city itself frames this as an early, actively-training phase rather than a rigorously evaluated productivity gain.

Implementation
Indicative cost
High (€500k–€5M) — No additional software licensing cost cited, since the rollout uses San Francisco's existing Microsoft enterprise agreement; other costs (training, change management) are not quantified.
Time to results
Medium (1–3 years) — Pilot ran approximately January-June 2025 (2,000+ staff); citywide rollout began July 2025 and continued through mid-2026, reaching ~30,000 employees across 40+ departments.
Staffing & skills
Delivered by the Office of the Mayor and Department of Technology with Microsoft and workforce-training partner InnovateUS, Paired with a five-week mandatory training programme for staff

Conditions for success

  • Used San Francisco's existing Microsoft enterprise licensing, so the rollout is described as carrying no additional software cost
  • Operated under the pre-existing 22J AI Transparency Ordinance, with updated Generative AI Guidelines (8 July 2026) requiring staff to take responsibility for AI-assisted material

Common failure modes

  • The headline productivity claim (up to 5 hours/week saved) comes from a self-reported pilot survey, not an independent time-and-motion study
  • No citywide before/after service-delivery metrics (e.g., case-processing times, error rates) have been published alongside the adoption numbers

Where it fits

Governance type
municipal government
Scale
citywide - ~30,000 employees, 40+ departments
Income level
high-income

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful