Cirrus Institute for AI and Data Governance

Success at the first Global Dialogue on AI Governance should be measured by whether the conversation meaningfully includes the people most affected by the decisions being made. From the Cirrus Institute's perspective -- where our work centers on the governance of emerging technologies and, specifically, on how AI displacement is reshaping labor markets across industries and experience level -- a successful Dialogue would accomplish three things. First, it would establish a shared taxonomy of AI risk that travels across jurisdictions. Fragmented regulatory regimes are not just an inconvenience, they are a governance fault line. When standards diverge, accountability falls through the gaps. A common framework for categorizing AI risk, even a minimal one, gives regulators, courts, and civil society a shared language to work from. Second, it would center workers, not just industries, in the conversation about labor displacement. The Future of Work is not an abstract policy category. It is nurses, attorneys, journalists, and analysts trying to understand what their professional lives look like in five years. Our Future of Work Initiative at Cirrus was built precisely because that question deserves rigorous, data-driven, publicly accessible answers rather than think-piece speculation. A successful Dialogue would commit to labor-facing research infrastructure as a governance priority, not an afterthought. Third, it would produce accountability mechanisms, not just principles. The history of technology governance is littered with well-worded frameworks that generated no institutional consequence. Success means identifying who is responsible for what when AI systems cause harm, and building the enforcement architecture to make that responsibility real. The Dialogue's first iteration should be judged by whether it lays groundwork that future institutions can actually build on.

• Safe, secure and trustworthy AI
• Social, economic, ethical, cultural, linguistic and technical implications of AI
• Interoperability of governance approaches
• Protection and promotion of human rights

My research on the future of work centers on how AI systems are reshaping labor markets, institutional governance, and economic sovereignty. Among the thematic areas identified in United Nations General Assembly Resolution 79/325, four priorities stand out as requiring urgent and sustained engagement. Safe, secure, and trustworthy AI: the foundation. Labor displacement, algorithmic management, and autonomous decision-making systems introduce systemic risks not only to workers but to economic stability. My work examines whether vulnerabilities in AI systems, particularly in agentic and distributed architectures, are inherent, requiring governance frameworks that embed safety at the design layer rather than relying solely on ex post regulation. Social, economic, ethical, cultural, linguistic, and technical implications: the effect. I focus on how AI reconfigures labor value, amplifies inequality across jurisdictions, and reshapes participation in digital economies, particularly in martinalized communities and the outsouring of labor from the US. Interoperability of governance approaches: need for cohesive regulations. Fragmented regulatory regimes risk regulatory arbitrage and uneven labor protections. I analyze how trade law, sanctions regimes, and digital governance frameworks can be aligned to prevent jurisdictional gaps while preserving innovation. Protection and promotion of human rights: the architecture. AI systems increasingly mediate access to work, wages, and opportunity, raising urgent concerns around discrimination, surveillance, and due process. Embedding rights-based safeguards into technical and legal architectures is a core priority of my work.

Two stand out, both relate to economic protections and freedoms - 1 - The financialization of work through digital assets and AI-integrated payment systems. The intersection of AI with blockchain, central bank digital currencies, and platform-based compensation models raises unresolved legal questions around jurisdiction, taxation, and worker protections. 2 - Economic sovereignty in AI-mediated labor markets. As AI systems and digital platforms concentrate control over data, compute, and labor intermediation, states risk losing regulatory and fiscal authority over their own workforces.

Governance gaps across safe, trustworthy AI, human rights protections, and interoperable regulation are already reshaping the U.S. labor market through an accelerated wave of AI-enabled outsourcing. Firms are no longer offshoring only routine back-office work; they are now fragmenting higher-skilled tasks, like legal research, software development, financial analysis, and creative production, into globally distributed, AI-mediated workflows. This affects industries ranging from legal services and finance to healthcare administration, media, and customer operations. The core challenge is that AI reduces coordination costs while obscuring accountability. U.S. companies can leverage global labor pools through platforms and AI tools without corresponding obligations under domestic labor, tax, or employment law frameworks. This creates downward pressure on wages, erodes job stability, and weakens bargaining power for domestic workers, while simultaneously exposing workers in lower-income jurisdictions to precarious, surveilled, and undercompensated digital labor conditions. The absence of interoperable governance standards allows firms to arbitrage regulatory differences across jurisdictions, effectively externalizing labor protections. Economically, this dynamic contributes to labor market polarization in the U.S.: high-skill workers who can leverage AI see productivity gains, while mid-tier knowledge workers face displacement or wage compression. At the same time, value generated through global digital labor chains is increasingly captured by a small number of platform and infrastructure providers, rather than distributed across workers or jurisdictions. However, there are meaningful opportunities. AI-enabled global labor integration could expand access to work, increase efficiency, and support new forms of cross-border collaboration if governed effectively. The U.S. is well-positioned to lead in establishing interoperable standards that align AI deployment with labor rights, transparency, and accountability. Developing enforceable frameworks (particularly around algorithmic management, data use, and worker classification) could mitigate exploitation while preserving the economic gains of a more connected, AI-driven workforce.

Importantly, the Dialogue can elevate implementation as a priority. While many initiatives articulate principles, fewer address how they are enforced across jurisdictions. By fostering coordination between legal, technical, and economic actors, the Dialogue can help translate high-level norms into practical, interoperable governance tools, including shared audit standards, certification schemes, and accountability mechanisms that operate across borders. Sharing tools, resources, and already implemented solutions/programming/ research is the biggest draw to us.

The AI Dialogue should build on and connect existing initiatives rather than reinvent governance frameworks. Key efforts include the OECD AI Principles or the UNESCO Recommendation on the Ethics of AI (working with them on a toolkit/co-depository for AI research).Technical standards bodies like ISO and IEEE also play a central role in operationalizing safety and interoperability. Equally important are academic and research institutions that are shaping cutting-edge legal and policy thinking. Centers close to the UN heeadquarters such as New York University, Cornell Tech, and the Knight First Amendment Institute at Columbia University are actively advancing scholarship on AI governance, including issues of speech, accountability, and platform power. These institutions play a critical role in bridging doctrinal legal analysis, particularly around First Amendment implications, and emerging AI regulatory challenges. Multistakeholder initiatives like the Global Partnership on Artificial Intelligence and development actors such as the World Bank (I'm attending their 2026 Youth Summit on future of work!) further contribute through capacity-building and applied policy experimentation. The added value of the AI Dialogue lies in its ability to connect these ecosystems, linking global norms, technical standards, and academic research into a more coherent governance architecture. It can identify gaps between principle and practice, particularly in areas like cross-border enforcement, AI-enabled labor markets, and rights-based safeguards.

• The AI Dialogue should be designed as a multi-layered, hybrid forum that moves beyond plenary statements toward sustained, outcome-oriented collaboration. Governments, industry, academia, civil society, and technical standard-setters should each contribute distinct competencies: states provide regulatory authority
• industry offers implementation insight
• academia contributes independent research
• and civil society ensures accountability and rights-based framing.

Smaller states and regulatory authorities with limited technical capacity are often excluded from meaningful agenda-setting, despite being significantly affected by AI-driven economic restructuring. Linguistic minorities and non-English-speaking stakeholders are also underrepresented, limiting the inclusivity of norm development. Inclusion requires more than formal participation. The Dialogue should provide financial and technical support for participation, enable remote and asynchronous engagement, and incorporate multilingual processes. Worker representation should be formalized through partnerships with labor organizations and digital worker collectives.

To move beyond static dialogue, the AI Dialogue could adopt interactive and experimental formats. First, policy simulation exercises ("AI war games") could allow stakeholders to test responses to scenarios such as labor displacement shocks, cross-border regulatory conflicts, or AI system failures. These simulations would surface governance gaps in real time. Second, live audit and red-teaming demonstrations, where technical experts evaluate AI systems against safety and rights benchmarks, could bridge principles with practice. Third, worker-centered forums (could include including anonymized testimony and case-based hearings) would ground discussions in lived experience, particularly for globally distributed digital labor forces. Finally, digital participation platforms (long-term community!) with transparent version control could allow continuous input beyond formal sessions, ensuring that the Dialogue remains adaptive, inclusive, and responsive.

AI Incident Database - https://incidentdatabase.ai/ ; https://airisk.mit.edu/ai-incident-tracker Academic Initiatives - Cornel Tech Digital Life Initiative (https://dli.tech.cornell.edu/), Knight Institute at Columbia Case Tracker (e.g., https://knightcolumbia.org/cases/amazon-v-perplexity-ai), National Conference of State Legislators (NCSL) AI legislative tracker (https://www.ncsl.org/financial-services/artificial-intelligence-legislation-database). Thinktanks that consistently put out quality resources or content: Globethics, Cambridge AI Safety Hub, ERA Cambridge, Institute for Law & AI, Collective Intelligence Project, CSET. Projects to Participate in: 2026 World Bank Group Youth Summit (I'm a Delegate), Aspen Institute Learning Initiatives, Civil Society Organizations and Academic Network on AI ethics convened by UNESCO, the Intellectual Property and Culture Subgroup, AI Impact Alliance (co-designing a global online repository to navigate the impact of AI on art, culture and knowledge creation, our institute is participating!). Formal Risk Frameworks - EU AI Act; ISO/IEC: 42001 (AI Management Systems)/5 (AI System Impact Assessments)/6 (AI System Audit and Certification) and associated ISO/IEC standards regarding AI System lifecycle processes (5338), guidance for AI Applications (5339), as well as bias (24027), risk management (23894), and data quality (5259); NIST AI Risk Management Framework (AI RMF); OECD Principles on AI.