A recent report from Code for America reveals that only a handful of states have established effective evaluation mechanisms to assess the public value of artificial intelligence (AI) deployments in government. The 2026 Government AI Landscape Assessment highlights that while there is a growing push for AI adoption among state governments, progress remains uneven, hindered by issues such as inadequate data quality and security concerns.
Released on Friday, the report marks a year since the first-ever assessment of government AI readiness. It categorizes states into four stages of AI maturity: early, developing, established, and advanced. The findings indicate that most states still find themselves at the early stages of their AI journey, which is framed through four primary phases: readiness, piloting, implementation, and impact.
The readiness phase focuses on building the necessary capabilities for AI implementation, while the piloting phase aims to identify effective strategies for scaling applications. The implementation stage seeks to create measurable outcomes, and the impact phase involves governance that promotes further innovation. This structured approach underscores the complexities involved in integrating AI into government operations.
Notably, the report reveals that nearly all states have initiated some form of AI pilot project, often utilizing generative AI tools. However, only a small fraction have successfully integrated AI into enterprise-scale workflows across multiple agencies. Alarmingly, just seven states were identified as having established comprehensive evaluation mechanisms to gauge the societal benefits of their AI initiatives.
The states that are leading in AI adoption tend to share common attributes, including robust executive leadership, effective cross-agency governance, and environments conducive to controlled experimentation. Structured pilot programs and enterprise-level data infrastructure also play crucial roles in these successful implementations. The report identifies Maryland, New Jersey, North Carolina, Pennsylvania, Texas, Utah, and Vermont as frontrunners in this evolving landscape.
The findings highlight a significant gap in the public sector’s ability to measure the effectiveness and value of AI initiatives, suggesting that many states may be moving forward without fully understanding the implications of their deployments. The lack of established frameworks to quantify outcomes means that the potential benefits of AI in government could remain largely unmeasured.
As AI continues to evolve, the urgency for proactive governance and accountability in its application grows. The 2026 Government AI Landscape Assessment serves as a critical reminder of the need for transparent evaluation processes that can foster public trust and ensure that AI contributes positively to society.
Looking ahead, it will be essential for states not only to ramp up their AI capabilities but also to develop metrics that align with public interests. As governments navigate the complexities of this technology, the stakes will be high for both operational efficiency and public confidence in AI’s role in governance.
See also
Independent Journalism Seeks Reader Support Amid Declining Ad Revenue
Anthropic Accuses Moonshot AI of 3.4M Unauthorized Claude Exchanges Amid US State Response
71% of Aussies Use Generative AI, Yet Only 36% Trust Its Implementation, Says Expert
US Defense Partners with Anthropic, OpenAI, and Tech Giants for AI-First Military Initiative
AI Technology Enhances Road Safety in U.S. Cities
