Loading...
Loading...
Colorado has been more deliberate about government AI strategy than most states its size, starting with Governor Polis's Executive Order 22-02 in February 2022, which directed the Colorado Office of Information Technology (OIT) to develop a state AI strategy, inventory existing AI uses, and establish governance principles before agencies deployed new systems. That order produced a living AI policy document that OIT has updated twice since — most recently in 2024 to address generative AI use by state employees, which had outpaced the original framework within 18 months of ChatGPT's release. Colorado's technology policy environment reflects a state that is simultaneously a major federal defense and intelligence hub (Space Command headquarters at Peterson Space Force Base, NORAD at Cheyenne Mountain, five major military installations in the Colorado Springs corridor) and a progressive governance laboratory where the Polis administration has been willing to move faster on citizen-facing digital services than most Mountain West peers. The Colorado Public Utilities Commission, which regulates the state's electric utilities, telecommunications carriers, and increasingly its transportation network companies, has become one of the more AI-forward state regulatory bodies — using ML-based rate case analysis tools that OIT developed as a shared service and that the PUC has customized for utility ratemaking proceedings. LocalAISource connects Colorado government clients with AI specialists who understand OIT's procurement frameworks, the Space Command defense tech ecosystem, and the specific demands of a state where ski season, wildfire season, and Front Range population growth create distinct demand cycles for government services.
Updated June 2026
Colorado's OIT serves all three branches of state government — a broader mandate than most state IT offices — and its AI governance framework reflects that breadth. The 2022 AI strategy identified six guiding principles (transparency, accountability, privacy, security, equity, and human oversight) and established a cross-agency AI Governance Council that meets quarterly and includes representatives from the Governor's Office, Department of Law, Colorado Civil Rights Division, and major cabinet agencies. This governance structure is not just ceremonial: the Governance Council has twice returned AI deployment proposals for additional equity review — once for a CDHS benefits eligibility tool and once for a CDPS criminal history check AI system — requests that added 4–6 months to those deployment timelines. OIT's shared technology services catalog includes an AI-as-a-service offering built on Microsoft Azure AI services, available to agencies through the statewide Azure Enterprise Agreement. Agencies with specific needs beyond that catalog work through OIT's IT Project Management framework, which requires a Project Initiation Document and Technology Acquisition Plan for any AI project above $150,000. The Technology Acquisition Plan must include a risk classification (OIT uses a three-tier model: low, significant, high-risk) and, for significant and high-risk systems, an AI impact assessment that addresses potential disparate impacts on protected classes. In practice, OIT's framework creates a 3–4 month overhead for first-time AI deployments in Colorado state government, but repeat deployments in the same agency with the same vendor run much faster because the governance artifacts are already established. Colorado's Department of Revenue, Department of Labor and Employment, and Colorado Department of Transportation have all completed multiple AI deployments and now run through OIT review in 6–8 weeks. For vendors entering the Colorado market for the first time, the shortlist criterion is: have you done the OIT documentation before? Agencies consistently prefer vendors who arrive with templated impact assessment documents and demonstrated OIT approval track records over those who treat governance as an afterthought.
Colorado Springs is home to five active military installations: Peterson Space Force Base (headquarters of Space Command since 2021), Schriever Space Force Base, Fort Carson, the Air Force Academy, and Cheyenne Mountain Space Force Station. The presence of Space Command headquarters — which relocated from California in 2021 in a politically contested move — has added a new layer of defense AI procurement to the Colorado Springs technology market. Space Command's requirements for AI in space domain awareness, satellite telemetry analysis, and orbital debris tracking represent a specialized AI market that Colorado-based defense contractors including Ball Aerospace (now BAE Systems), General Dynamics IT, and Parsons have been active in developing. The spillover from defense AI into civilian Colorado government is real but uneven. Colorado Springs city government, which hosts more active-duty military population than any Colorado municipality, has benefited from defense-tech talent availability — the city's IT department has recruited several former Space Command data scientists who now run civilian AI programs. The Colorado Cybersecurity Center, a state-funded partnership between Colorado Springs-based defense contractors and OIT, provides threat intelligence and AI-powered security monitoring to state agencies and local governments that lack their own SOC capabilities. At the state level, the Colorado Department of Military and Veterans Affairs has used defense-sector relationships to accelerate several AI procurement decisions — particularly in case management tools for veterans benefits navigation. The department's AI-assisted benefits eligibility checker, which helps veterans identify state and federal programs they qualify for based on service records and current circumstances, was developed with support from Space Command's AI Accelerator program, which includes a civilian government technology transfer component. The Governor's Innovation Office has specifically cited the defense-civilian technology spillover in the I-25 corridor as a model it wants to replicate for other sectors.
The Colorado Public Utilities Commission regulates electric utilities, natural gas, telecommunications, and transportation network companies — and its docket management workload has grown substantially as the Polis administration pursues aggressive clean energy transition policies that generate contested utility ratemaking proceedings. The PUC's AI deployment is a useful case study in how regulatory bodies can use AI to reduce the analytical burden without automating decisions that require human judgment. OIT developed an NLP-based docket analysis tool specifically for the PUC that classifies filings, extracts key claims and financial data, and generates staff briefing summaries — reducing the time from filing to staff memo from 12 days to 4 days on standard rate cases. The PUC's commissioners have explicitly noted that the tool improves staff bandwidth without changing the human deliberation process. Colorado's Department of Human Services manages SNAP, Medicaid support, child welfare, and adult protective services for a state that has seen significant population growth — Front Range counties have grown 15–20% since 2015, creating sustained caseload increases that staffing has not kept pace with. CDHS deployed an ML-based case risk stratification tool for child welfare in 2023, built on the Predictive Risk Modeling framework that Colorado had previously piloted (and then paused following a 2018 equity audit that found the original model had training data biases). The 2023 version included independent equity testing through the University of Denver's Institute for Human Development and Family Studies before deployment, which allowed the OIT Governance Council to approve it as a significant-risk system rather than requiring high-risk review. Colorado's wildfire season creates a government AI demand pattern that is genuinely state-specific. CDPS's Division of Fire Prevention and Control uses ML-based fire behavior modeling that cross-references weather data from NOAA, terrain models from USGS, and fuel-load data from the Colorado State Forest Service — the system provides real-time spread probability modeling during active fires and has been credited with improving evacuation order accuracy (earlier evacuations in areas where fires spread faster than expected, avoiding premature orders in areas that models showed lower risk). That model runs on OIT's cloud infrastructure and is shared with county emergency management offices through the state's Enterprise Geographic Information System.
Strategic planning for AI adoption, readiness assessment, and roadmap development
Workflow automation using AI, including Make.com-style automation and RPA
Predictive models, data analysis, and ML pipeline development
Text analysis, document automation, sentiment analysis, and language processing
EO 22-02 directed OIT to develop an AI strategy, conduct an inventory of existing state AI uses, and establish governance principles within 120 days. OIT delivered those in October 2022. The 2024 update added specific provisions for generative AI: state employees may use approved generative AI tools (primarily Microsoft Copilot through the state's M365 agreement) but are prohibited from inputting personal information, confidential business information, or attorney-client privileged content into AI tools without explicit data use authorization. The update also required agencies to disclose AI use in communications to the public when AI generated or substantially assisted the content — a transparency requirement that Colorado is among the first states to implement at the executive branch level.
The PUC's docket analysis tool classifies filings, extracts financial data and key claims, and generates staff briefing summaries that reduce time-to-memo from 12 to 4 days. The tool does not generate recommendations or draft orders — it prepares information for human decision-makers. The key design choice that got it through OIT's governance review was strict scoping: the AI handles information organization, not analysis, and every output is reviewed by a staff attorney before use. For other state regulatory bodies (State Banking Commission, Division of Insurance), this scoped-AI model is the path most likely to clear both legal and governance review — AI that accelerates staff work without substituting for expert judgment.
The relationship is primarily talent and technology transfer. Space Command and the broader Colorado Springs defense complex employ thousands of data scientists and ML engineers whose skills are directly applicable to civilian government AI. Colorado Springs city government and several El Paso County agencies have recruited former defense tech staff. The Colorado Cybersecurity Center, funded by a state-defense partnership, provides AI-powered threat monitoring to local governments that can't afford commercial SOC services. The practical channel for accessing defense-sector AI expertise for civilian government work is through Colorado-based defense contractors (Ball Aerospace/BAE Systems, Parsons, SAIC Colorado Springs) that have built civilian government practice areas alongside their defense work.
For agencies using OIT's AI-as-a-service catalog (Azure Cognitive Services, Power Automate with AI Builder), per-use costs are baked into the state's enterprise Azure agreement — agencies pay for consumption but not additional procurement overhead. Custom AI projects — model development, training data preparation, algorithmic impact assessments — typically run $80,000–$350,000 for a first deployment in a state agency context, with OIT's governance review adding approximately $25,000–$75,000 in documentation and assessment work. Federal matching funds are available for AI in Medicaid (90/10 DHCS), workforce programs (50/50 CDLE), and transportation (80/20 CDOT/FHWA). Colorado also has a Governor's Office of Innovation grants program that has funded several AI pilots in smaller agencies.
Colorado experiences 1,500–2,000 wildfire starts annually, with catastrophic fire seasons (2020, 2021, 2022) producing multi-billion-dollar economic losses and statewide emergency declarations. CDPS's fire behavior modeling system is the highest-profile AI deployment, but the demand extends to CDOT traffic management (real-time I-70 corridor capacity routing when evacuations are ordered), CDPS dispatch optimization (routing mutual-aid crews from across the state), and CDLE labor market modeling (short-term economic impact estimates for affected counties used in SBA disaster declaration requests). Vendors who have experience with wildfire-context AI — particularly those who have worked with CAL FIRE, Oregon Department of Forestry, or USFS — have a significant credibility advantage in Colorado procurement.
Get found by Colorado businesses on LocalAISource.