As federal AI regulation stalls in Washington, state and regional leaders are stepping into the void with their own governance frameworks, prioritizing worker protections, community oversight, and transparent decision-making over rapid tech deployment. Maryland Governor Wes Moore signed an executive order establishing AI principles and data center regulations, while the Common Market for Eastern and Southern Africa (COMESA) is developing a regional AI strategy and model regulatory framework. Together, these initiatives signal a fundamental shift in how AI governance is being approached globally.

Why Are States and Regions Acting When Congress Won't?

Congress has failed to pass comprehensive AI legislation despite mounting concerns from voters and industry experts alike. This inaction has created a governance vacuum that state and regional leaders are now filling. Governor Moore made this frustration explicit, stating that "AI is moving faster than the safeguards around it. Marylanders should not have to wait for a disaster or, frankly, for Congress to be protected". His sentiment reflects a broader recognition that waiting for federal action is no longer viable when AI systems are already operating in communities and workplaces.

The Maryland approach is particularly notable because it addresses not just AI systems themselves, but the infrastructure supporting them. The executive order creates a unified review process for large data centers, which consume enormous amounts of electricity and water. Moore plans to push the state legislature to repeal tax exemptions for data centers, a move that environmental groups and local government associations have praised as overdue accountability.

What Specific Protections Are These Frameworks Building?

Maryland's AI framework centers on five core principles designed to protect residents, workers, and children. These principles address multiple dimensions of AI's impact on society, moving beyond the narrow focus on model safety that dominates tech industry discussions.

• Worker Protections: The framework calls for greater involvement from employees and unions as AI is introduced into workplaces, along with additional training and programs to help workers adapt to automation and AI-assisted roles.
• Child Safety: The plan mandates stronger protections for children using AI chatbots and social media, including standards for AI tools used in Maryland schools and increased protections against addictive design patterns.
• Accountability in High-Risk Decisions: AI systems cannot "quietly decide who gets a loan, or housing, or healthcare or a government benefit without accountability and a human in the loop," according to Governor Moore, reflecting a commitment to human oversight in consequential decisions.
• Community Voice: Local residents living near proposed data center sites will have greater influence in the approval process, ensuring that infrastructure decisions reflect community priorities rather than developer timelines.
• Transparency and Accountability: The framework includes provisions for independent third-party audits, whistleblower protections, and a right of publicity statute to prevent unauthorized use of individuals' likenesses in AI training data.

Meanwhile, COMESA is taking a different but complementary approach by building regional consensus across 21 member states in Eastern and Southern Africa. The organization is developing a Regional AI Strategy, Model Policy Guidelines, and Model Regulatory Framework based on validated evidence from across the region. This approach recognizes that AI governance cannot be one-size-fits-all; it must account for regional economic conditions, infrastructure capacity, and development priorities.

How Are These Frameworks Being Developed?

Both initiatives emphasize inclusive, evidence-based development rather than top-down mandates. Maryland's approach involves an AI subcabinet developing legislative recommendations that will be presented to the General Assembly, ensuring that policy reflects input from multiple state agencies and stakeholders. This collaborative process helps build political support and ensures that regulations account for practical implementation challenges.

COMESA's process is even more expansive. The organization convened a Regional Validation Meeting in Eswatini from September 16 to 18, 2026, bringing together member states, regional economic communities, development partners, and AI experts to validate evidence and identify gaps in the regional AI ecosystem. This three-day meeting was designed to ensure that the final policy instruments "rest on an accurate, credible and shared understanding of the region's AI landscape," according to Eswatini's Acting Minister of Information, Communications and Technology.

"AI can accelerate development and improve well-being across COMESA, but this requires investment in skills, infrastructure and data, alongside safeguards for privacy, security, fairness and accountability," stated Leonard Chitundu, COMESA Telecommunications Officer.

Leonard Chitundu, COMESA Telecommunications Officer

The COMESA validation process highlights a critical insight: governance frameworks are only effective if they reflect the actual conditions and priorities of the communities they're meant to protect. By involving member states in validating the evidence base, COMESA is building ownership and ensuring that the resulting policies will be implementable across diverse economic and technological contexts.

What Does This Mean for the Future of AI Governance?

These parallel initiatives suggest that AI governance is becoming decentralized and place-based rather than centralized and universal. Maryland's focus on data center regulation and worker protections reflects the concerns of an industrial state with significant tech infrastructure. COMESA's emphasis on skills development, infrastructure investment, and inclusive digital transformation reflects the priorities of developing economies seeking to harness AI's benefits while avoiding the pitfalls that wealthier nations have experienced.

Governor Moore articulated a nuanced view of the competitive and cooperative dimensions of this emerging landscape. "My fear is not China. My fear is a technology that no human being can control. And so we've got to understand that there has to be that delicate balance of cooperation and competition," he stated. This framing suggests that state and regional leaders recognize AI governance as a shared challenge requiring both healthy competition between jurisdictions and cooperation on fundamental safety principles.

Governor Moore

The Maryland and COMESA initiatives also demonstrate that AI governance is not purely a technical or regulatory problem; it's fundamentally about power and accountability. By requiring community input, worker involvement, and transparent decision-making processes, these frameworks are asserting that AI deployment decisions should not be made unilaterally by technology companies or distant federal agencies. Instead, they should reflect the values and priorities of the communities most affected by these systems.

As more states and regions develop their own AI governance frameworks, the question of interoperability and consistency will become increasingly important. Companies operating across multiple jurisdictions will need to navigate different regulatory requirements, which could either drive convergence around best practices or create fragmentation that slows innovation. The coming months will reveal whether these emerging frameworks inspire federal action or whether the United States will continue to rely on a patchwork of state and local regulations while other regions move toward more coordinated approaches.