Colorado Appoints Former Federal AI Chief Jane Yang to Lead State AI Architecture

What Happened

The Colorado Governor’s Office of Information Technology has appointed Jane Yang as principal director for AI architecture. Yang joins from the federal Administration for Children and Families, where she served as chief artificial intelligence officer from November 2024 through June 2026. She officially joined Colorado OIT this month.

Her background spans federal AI leadership, enterprise data strategy, privacy-focused technology companies, and humanitarian organizations including the International Rescue Committee and One Acre Fund. She also held roles at Basecamp and Deloitte in data and research capacities.

Why This Hire Is Significant

Yang’s federal experience is directly relevant to what Colorado needs. At ACF, she led the launch of a model-agnostic enterprise generative AI platform that allowed teams to build AI assistants tailored to their specific workflows, while maintaining secure integration of large language model capabilities into custom applications.

She also built ACF’s first enterprise data strategy and ran an internal program called “AI street sense,” designed to help employees develop practical judgment about when AI is genuinely useful and when it introduces more risk than value. That kind of operational literacy is rare and difficult to scale.

The Broader Context: OIT Restructuring

This appointment does not happen in isolation. Colorado OIT is in the middle of a significant organizational restructuring, triggered by an internal operational assessment that identified serious shortcomings across the department. The restructuring led to the departure of former CIO David Edinger, the elimination of 173 positions, and a transition to a new operating model.

Sarah Tuneberg, previously deputy executive director for digital and delivery, has stepped into the top technology role. Yang’s hire, alongside the recent appointment of Casey Cook as principal director of enterprise architecture, represents one of the first concrete leadership moves under the new structure.

The Governance Tension Yang Has Named Directly

Yang has been explicit about the challenge ahead. In a recent LinkedIn post, she acknowledged that AI can meaningfully improve how Coloradans access public services—faster information retrieval, streamlined benefits applications, reduced administrative friction—while also introducing cybersecurity vulnerabilities and real risks of harm if deployed without proper disclosure, evaluation, and appeal mechanisms.

Her framing is worth noting: she described the challenge not just as a technical one, but as a trust-rebuilding exercise with agency partners, oversight bodies, and the public. That is a more honest and more difficult mandate than simply deploying AI tools.

What This Means for the AI Tools Ecosystem

For vendors and builders working in the public-sector AI space, Colorado’s direction under Yang points toward a few clear priorities:

• Model-agnostic infrastructure over single-vendor lock-in
• Secure LLM integration with governance guardrails built in from the start
• Enterprise data strategy as a prerequisite, not an afterthought
• Explainability and appeal pathways as non-negotiable deployment requirements

State governments are becoming more sophisticated buyers of AI infrastructure. Appointments like this one indicate that the era of selling AI tools to government on the basis of capability alone is narrowing. Governance architecture, security posture, and operational accountability are increasingly part of the evaluation criteria.

Colorado is not the only state moving in this direction, but it now has someone in a key role who has built this kind of infrastructure at the federal level before. That is a meaningful difference from hiring a generalist.