The Real AI Tech Stack: Local to Global Governance

On May 7-9, The Special Competitive Studies Project held its annual AI+ Expo, an event that has become widely recognized in D.C. for the depth and breadth of companies, universities, government officials, and diplomats engaged in the discussion and practice of AI. 

Beneath the flashy drone shows, walking robots, and multitudes of AI companies from startups to industry giants scattered throughout the expo hall, there was a clear thematic undercurrent.

AI is often described as a “stack,” spanning software to hardware and everything in between. At the expo, it became clear that AI governance also plays out as its own stack – the layers being local and state level policies, national strategies, and global approaches.

The Local Layer

The foundation of this stack – the local and state level – was evident in the presence of state and local level leaders at the summit. On a central stage, Mississippi Governor Tate Reeves spoke about how state leadership is helping position the U.S. to compete globally in AI.

Universities were equally central to this local layer. From Texas to California to D.C. and beyond, academic institutions lined the dedicated AI education spaces, showcasing their unique AI degree programs, research centers, and startup ecosystems.  

These institutions were consistently cited by several speakers throughout the conference as essential in shaping the future of AI. For example, SEMI, the industry association representing the semiconductor sector, pointed to the “brain drain” of government professionals in D.C. that has led to organizations like theirs relying on policy research coming out of think tanks, nonprofits, and universities to inform their members on trends in AI and advanced manufacturing. 

The influence of local approaches to AI is not lost on industry leaders, although there is an active debate on the pros and cons. At an event held by the Center for International Private Enterprise, the International Republican Institute, and The National Endowment for Democracy, some private sector panelists cautioned against the “over 1700” AI-related local laws and ordinances across the country, which they claim may lead to stifled innovation and an untenable regulatory landscape. 

At the same time, Pennsylvania is offering a notable example of how subnational governance can work in practice. In late 2025, the University of Pennsylvania signed a cooperative agreement with the Commonwealth of Pennsylvania to provide AI advisory support. Leveraging Penn AI’s interdisciplinary expertise across education, ethics and governance, the partnership is helping the state responsibly navigate AI policy, organizational strategy, governance, and risk assessment. This effort also integrates perspectives from the public sector and industry. 

Pulling from the expertise of civil society was an approach suggested by both U.S. and international panelists. While some disagreed on whether fragmented governance would create long-term consequences, there was consensus that any effective AI policy at any level must be informed by a breadth of stakeholders, from policymakers, to industry leaders, to academic experts, and users; an approach that the University of Pennsylvania is pioneering. 

The Global Layer

If local and state approaches form the foundation of the AI Governance stack, and build to a national strategy, the next layer is how those efforts play out on a global stage. Albert S. Yohai from the Colombian Chamber of Information and Technology reframed the widely discussed “AI Race,” noting that the real challenge for many emerging democracies is not figuring out how to outpace authoritarian nations to dominate the values and governance associated with AI, but figuring out how to get to the starting line in the first place. This raised a fundamental question discussed by the panel: how do countries develop the workforce, infrastructure, and capacity to engage meaningfully in both the AI tech and AI governance stacks? 

U.S.-led initiatives like Pax Silica highlight the importance of this question. On the panel, Heather Goethert of the U.S. Department of State described this global economic security coalition as an effort to promote innovation-driven supply chains in AI, semiconductors, critical minerals, and more. 

The participating countries (so far) largely represent democracies with shared values around technology standards, but also ones that have the capability to immediately contribute to the coalition. 

That “capability” isn’t a given for many emerging economies, like Columbia, that may seek to contribute and benefit from such a coalition. That’s where private sector and university engagement plays a meaningful role. Similar to initiatives held under Penn AI, panelists from Google and SEMI spoke to the workforce development efforts in the U.S. and around the world that are upskilling human capital in AI literacy – so more countries can participate in the AI ecosystem and contribute to questions of global governance. 

If the AI Tech Stack and the AI Governance Stack mirror each other, workforce development is what connects every layer – enabling local innovation and coordination across industries and borders. What emerged from the depths of the AI Expo was not just a flashy showcase of cutting–edge technology, but a recognition of the interconnected layers shaping how AI is governed in an increasingly complex domestic and geopolitical environment. 

Nikki Hinshaw is a Global Policy Programs Coordinator at Penn Washington