Colorado’s first principal director of AI architecture, Jane Yang, is shaping the state’s AI direction as the Office of Information Technology (OIT) carries out a large-scale realignment. Yang joined OIT in July with federal government and private-sector experience. The state is moving to a product-oriented delivery model to better support agencies, and Yang said change management is central to both the AI work and OIT’s operating reset.

Yang said the state’s AI directive has three pillars: accelerating delivery, activating opportunities and maintaining transparency. She calls the judgment needed to decide when to use AI, select an appropriate tool and disclose its use “AI street sense.” The approach also includes giving employees access to secure AI tools aligned with state governance. Colorado plans to open source as much AI guidance and other resources as possible so other states can use them, and Yang said Colorado wants to contribute as states collaborate on digital government services.

Yang’s goal is to make AI part of technology practices across state agencies rather than a specialized capability held by a small team. Colorado is preparing for Senate Bill 26-189, House Bill 26-1263 and House Bill 26-1139, which take effect in 2027. The central AI team is term-limited and is planned to sunset within three to five years; within five years, Yang hopes OIT’s ongoing functions will incorporate the AI vision and the team will shift toward advising. She said the aim is to avoid a permanent AI bureaucracy while building documented processes, tools, best practices and infrastructure for teams to continually assess large language model products. Officials are expected to evaluate whether AI delivers intended outcomes and target returns on investment.

Colorado’s first principal director of AI architecture, Jane Yang, is shaping the state’s AI direction as the Office of Information Technology (OIT) carries out a large-scale realignment. Yang joined OIT in July with federal government and private-sector experience. The state is moving to a product-oriented delivery model to better support agencies, and Yang said change management is central to both the AI work and OIT’s operating reset.

Yang said the state’s AI directive has three pillars: accelerating delivery, activating opportunities and maintaining transparency. She calls the judgment needed to decide when to use AI, select an appropriate tool and disclose its use “AI street sense.” The approach also includes giving employees access to secure AI tools aligned with state governance. Colorado plans to open source as much AI guidance and other resources as possible so other states can use them, and Yang said Colorado wants to contribute as states collaborate on digital government services.

Yang’s goal is to make AI part of technology practices across state agencies rather than a specialized capability held by a small team. Colorado is preparing for Senate Bill 26-189, House Bill 26-1263 and House Bill 26-1139, which take effect in 2027. The central AI team is term-limited and is planned to sunset within three to five years; within five years, Yang hopes OIT’s ongoing functions will incorporate the AI vision and the team will shift toward advising. She said the aim is to avoid a permanent AI bureaucracy while building documented processes, tools, best practices and infrastructure for teams to continually assess large language model products. Officials are expected to evaluate whether AI delivers intended outcomes and target returns on investment.