From the SLED-wide archive edition of August 29, 2026
State workforce pilot reports large perceived savings but uneven readiness
Commonwealth of Pennsylvania · State government · Pennsylvania, United States
- Publisher
- Lessons from Pennsylvania's Generative AI Pilot with ChatGPT
- Original publication
- March 2025
- Source retrieved
- Not recorded in the historical archive
What happened
Pennsylvania equipped 175 employees across 14 agencies with ChatGPT Enterprise for a yearlong pilot using surveys, focus groups, office hours, and role-specific support.
Why it matters
The pilot offers a state-government operating model for exploring broad knowledge-work use while documenting the adoption barriers that can keep a nominally available tool from becoming routine practice.
Evidence and measured results
Participants estimated saving 95 minutes per day and most described the experience as very positive. The report also found no single successful-user profile and documented inaccuracy, habit formation, lack of learning time, a steep learning curve, and privacy uncertainty as material barriers.
Limitations and uncertainty
The evaluation was a volunteer pilot, not a controlled study; the 95-minute estimate was self-reported, 136 of 175 participants provided direct feedback, and Carnegie Mellon supported the effort consultatively rather than acting as an independent evaluator.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source as summarized in the preserved archive. Enriched 2026-09-05; this does not change the original publication date. Labels below come from the analysis itself.
Sales
Role takeaway
- Customer problem
- agencies may provide an enterprise assistant while staff remain unsure how to use it safely or make it part of routine work.
- Stakeholders
- innovation and program leaders, workforce development, IT, privacy, managers, and prospective ambassadors.
- Discovery
- which tasks recur; what learning time is available; where do employees misunderstand input processing; and who reviews generated work?
- Value hypothesis
- role-specific support may improve useful adoption and reveal measurable workflow gains.
- Potential engagement
- a use-case and readiness workshop followed by supported pilot cohorts.
- Unsupported claims
- participants' estimated 95 minutes saved per day is not a validated forecast, and the volunteer pilot does not establish causal productivity gains or independent evaluation by Carnegie Mellon.
Pre-sales engineering
Role takeaway
- Fit
- begin with approved knowledge-work tasks in an enterprise environment, retaining human review of outputs.
- Architecture and integration
- pair access with a role-based use-case library, scenario-specific safe inputs, and workflow telemetry; do not assume deeper integrations are required.
- Prerequisites
- confirmed input-processing and retention terms, accessible guidance, baseline tasks, and an owner for output quality.
- Constraints
- inaccuracies, habit formation, and learning demands can limit value despite positive sentiment.
- Security
- test the approved handling rules against realistic sensitive-input scenarios and explain where prompts and outputs are stored or processed. Proposed proof: compare observable cycle time and reviewed quality with participant estimates across different roles, tracking unsafe-input confusion and usability barriers separately.
Delivery
Role takeaway
- Work
- recruit agency ambassadors, run office hours and communities of practice, provide practice time, and maintain a reviewed use-case library.
- Dependencies
- manager support, privacy answers in plain language, and enough baseline evidence to assess actual workflow change.
- Ownership
- program managers own work products; an adoption lead coordinates support; IT/privacy teams maintain environment and handling guidance.
- Skills and adoption
- teach task selection, output checking, and safe input through role-specific examples instead of relying on a generic policy.
- Governance checkpoints
- onboarding readiness, cohort feedback, and review before higher-risk or team-based use.
- Proposed acceptance
- participants demonstrate approved handling, use targeted workflows routinely, and document observed quality/time changes. Risks include self-report inflation, inadequate learning time, and persistent privacy uncertainty despite enterprise terms.
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Provide an approved enterprise environment, but pair it with a role-based use-case library, safe input examples, output-review steps, and telemetry that can test self-reported savings against observable workflow measures.
Governance
Who approves, reviews and stays accountable for outcomes?
Use embedded AI ambassadors, communities of practice, simple dos and don'ts, human ownership of work products, and repeated user research before scaling to higher-risk or team-based workflows.
Security and privacy
What data, permissions and controls need testing?
Participants remained uncertain about how inputs were processed and stored despite enterprise terms and existing policy, showing that contractual protection must be translated into plain, scenario-specific guidance.
The preserved archive analysis covered architecture, governance and security. Not assessed for this record: accessibility and workforce, procurement, operating model.
Publication history
- 2026-08-29SLED-wide archive · Issue 0214 resources
Stable resource ID: pennsylvania-chatgpt-pilot