The national AI framework changes how state and local governments are expected to use AI. Over the past few years, many agencies have tried AI in limited ways, such as chat interfaces, document summarization, or internal support tools. These efforts were often small and managed within individual teams. The framework now requires agencies to take a broader view and define how AI will be used across the entire organization.

This introduces new responsibilities for leadership. AI is no longer treated as a technical add-on. It directly affects how services are delivered, how decisions are made, and how public data is handled. Agencies need to clearly define what types of data can be used, what level of human oversight is required, and how outcomes are reviewed. Without these rules, expanding the use of AI can lead to inconsistent decisions and unclear accountability.

Procurement becomes a central part of this shift. Traditional software can often be evaluated based on features, price, and integration. AI systems are different. Agencies need to understand how models are trained, what data they rely on, and how they behave under different conditions. This includes assessing risks such as bias, incorrect outputs, and dependence on external data sources. Contracts may also need to include requirements around transparency, updates, and performance monitoring.

The structure of state IT systems plays a large role in how easily these changes can be implemented. In states where technology decisions are controlled centrally, it is easier to apply common standards and review processes. In states where departments operate more independently, leaders need to rely on shared guidelines and coordination between teams. Hybrid approaches require both clear rules and flexibility, which can be harder to maintain over time.

Coordination across teams becomes one of the main challenges. AI use touches multiple groups, including IT, legal, procurement, and program teams. Each group has a different role in selecting, approving, and using AI. If these groups do not work together, agencies may end up using different tools for similar tasks, creating duplication and increasing risk. The framework encourages leaders to align these groups so that decisions are made consistently.

Another key point is ongoing oversight. AI systems are not static. Their performance can change based on new data, vendor updates, or shifts in how they are used. Agencies need processes to monitor outputs, review decisions, and make adjustments when needed. This requires regular checks rather than one-time approvals.

The agencies that move carefully and define clear rules, review processes, and coordination methods will be able to expand AI use with fewer issues. Those that move quickly without structure may face problems related to inconsistent decisions, unclear accountability, and vendor dependence.

Build the right capabilities for what’s coming. Book a quick call with the Fractional Source team.

In an episode of Fort Worth Innovates, host Cameron Cushman speaks with Chris Naler of General Genomics about how the company applies AI and machine learning to large healthcare and genomic datasets to identify disease risks earlier, particularly for conditions like Alzheimer’s and dementia, while also using similar methods in national security contexts for pattern detection and risk analysis. The conversation positions Fort Worth as a growing hub for applied AI, alongside efforts such as its executive workshop aimed at helping leaders understand how to adopt these technologies.

Harvard technology leaders and campus IT teams will meet at Memorial Hall in Cambridge to discuss how digital services, infrastructure, security, and academic tools can better support the university. The program includes main sessions, smaller breakouts, exhibits, and a closing reception, with participation from Harvardʼs CIO Council and leaders across schools and central IT. Details →

This summit gathers state and local leaders to explore how new technology can make government work better. It covers topics like cybersecurity, AI, data management, and digital services. Sessions are built to encourage teamwork, share useful advice, and help agencies move from planning to action. The event is free for public-sector participants and features a keynote from Ted Ross, Chief Information Officer for the City of Los Angeles. Details →

Agencies are encouraged to train staff in related areas such as AI, data, cloud, and security. This helps teams work together better, deliver faster, and stay flexible as technology blends. Instead of only hiring new people, they can use skills checks, tailored learning paths, and hands-on rotations to build a culture of continuous learning. Read →

Local governments are being encouraged to look at the full cost of owning roads, buildings, and services. If they only focus on upfront spending and ignore longterm upkeep, assets break down and hidden costs pile up. By tracking true costs over time, leaders can budget smarter, compare programs by value for money, and shift resources to where they have the greatest impact. Arlington County demonstrates this by using cost data to align staffing and services with actual community needs.