The AI Governance Gap: Why Global Agreements Mean Little Without Corporate Action

Authored by Carmen Skipworth, AI Strategist| Author | Speaker| Contributor to LA Tribune and Forbes Business Council |Global Women Peace Ambassador| Podcast Host of AI Made Simple with Carmen

Artificial intelligence is advancing faster than the institutions responsible for governing it. Governments are negotiating international agreements, policymakers are drafting regulatory frameworks, and global organizations are calling for responsible development.

Meanwhile, businesses are already deploying AI systems that make decisions, influence employees, serve customers, and increasingly act without direct human supervision. This raises a question that deserves far more attention: what good are international AI agreements if the organizations implementing the technology are not prepared to govern it?

Having participated in discussions on AI governance at the United Nations, I have seen a growing recognition that artificial intelligence is no longer simply a technology issue. It is an economic, social, ethical, and increasingly geopolitical challenge. Yet I believe one of the greatest risks we face is not the absence of international cooperation. It is the widening gap between the principles governments endorse and the decisions companies make every day.

From Global Principles to Corporate Practice

Over the past several years, international organizations have made meaningful progress toward shared expectations for AI. The United Nations has elevated AI governance within its broader discussions of global digital cooperation, UNESCO has adopted its Recommendation on the Ethics of Artificial Intelligence, and the European Union has established a comprehensive legal framework through the AI Act. These developments demonstrate that governments increasingly recognize the need for transparency, accountability, fairness, and human oversight.

However, an international declaration does not automatically change how an AI system operates inside a company. A commitment to transparency means little if employees cannot explain how an automated decision was reached, and a promise of accountability becomes questionable when no one is clearly responsible for the consequences of an AI-generated recommendation. The central challenge is translating high-level principles into operational decisions, and that responsibility cannot rest with governments alone.

AI Governance Is a Leadership Responsibility

For years, companies treated artificial intelligence primarily as an IT initiative, with technology departments evaluating software, implementing systems, and managing technical risks. That approach is increasingly inadequate. AI now influences recruitment, financial analysis, healthcare administration, customer service, cybersecurity, and strategic planning, and as AI agents become more capable of
executing multistep workflows, organizations must confront questions that extend well beyond technical performance. Who is accountable when an AI system makes a harmful decision? What information should an autonomous agent be allowed to access? When must a human intervene? How does a company recognize that its systems are behaving unexpectedly?

These are leadership and governance questions, not simply engineering challenges. Boards and executive teams must establish clear lines of accountability, define acceptable uses of AI, assess risks before deployment, and ensure that employees understand both the capabilities and the limitations of the systems they use. Governance should not be viewed as an obstacle to innovation. Properly designed, it provides the foundation for sustainable innovation by helping organizations build trust with customers, employees, regulators, and investors.

The Rise of AI Agents Raises the Stakes

The emergence of agentic AI makes corporate accountability even more urgent. Traditional AI systems primarily generated information or recommendations. Increasingly, AI agents can interact with software, retrieve sensitive information, communicate with external parties, and execute business processes with varying degrees of autonomy. This creates extraordinary opportunities for productivity and efficiency, but it also introduces new operational risks.

Consider an AI agent responsible for managing customer inquiries. Well designed, it can improve response times, identify opportunities, and reduce administrative burdens. But what happens if that same agent accesses information beyond its intended scope, makes inaccurate financial commitments, or takes actions that violate company policy? An organization cannot simply point to its technology provider and claim the system was responsible. Businesses remain accountable for how they deploy and supervise AI. That is why companies need clearly defined operating boundaries, access controls, monitoring systems, escalation procedures, and the ability to suspend AI activity when necessary. The more independently AI systems can act, the more important meaningful human oversight becomes.

Global Governance Must Include More Than the Largest Economies

Another concern is whether emerging international frameworks adequately reflect the needs of countries outside the major technology powers. Much of the global conversation is shaped by the United States, the European Union, and China, yet the consequences of AI adoption will extend far beyond these economies. In Southeast Asia, for example, countries are pursuing AI-driven economic growth while confronting challenges in digital infrastructure, workforce readiness, education, cybersecurity, and regulatory capacity. A governance framework designed for a multinational technology company may not translate easily to a small enterprise in Thailand, Laos, or Vietnam.

If international cooperation is to be meaningful, it must account for differences in economic development, institutional capacity, language, and access to technology. It must also ensure that women, smaller businesses, and historically underrepresented communities have a voice in shaping the future of AI.

Inclusion is not merely an ethical aspiration; it is essential to developing governance approaches that work across different societies.

AI Governance and Global Peace

As a Global Women Peace Ambassador, I believe AI governance is also closely connected to peace and international stability. Artificial intelligence can support scientific collaboration, humanitarian assistance, education, and economic opportunity. At the same time, poorly governed systems can amplify misinformation, deepen social divisions, reinforce discrimination, and contribute to geopolitical instability.

Responsible AI development therefore requires more than preventing technical failures. It requires protecting human dignity, strengthening public trust, and ensuring that technological progress does not undermine social cohesion. This matters even more as AI becomes embedded in the information systems and institutions that shape public understanding and decision-making. Peace in the digital age will depend, in part, on our ability to build technology that supports cooperation rather than division.

Closing the Gap Between Commitment and Action

The next phase of AI governance should focus less on producing additional declarations and more on demonstrating measurable accountability. International agreements can establish shared expectations, but businesses must translate those expectations into policies, training, risk assessments, and everyday operating procedures. Executives should be able to explain which AI systems their organizations use, what decisions those systems influence, who oversees them, and how failures are identified and addressed.

Governments, for their part, must develop practical guidance and enforceable requirements suited to different levels of risk, while supporting businesses that lack extensive technical and compliance resources. Universities and educational institutions also have a role to play in preparing future leaders to understand AI governance as a core management responsibility.

Ultimately, the success of global AI governance will not be measured by the number of agreements signed or conferences held. It will be measured by whether artificial intelligence operates in ways that are safe, accountable, inclusive, and beneficial to society. We do not have to choose between innovation and responsibility; the real challenge is ensuring they advance together. The future of AI will not be determined by technology alone. It will be shaped by the leadership decisions we make today.

About the Author

Carmen Skipworth is an AI leadership strategist, international speaker, and author of Lead with Intelligence: A Woman’s Guide to Thriving in the Age of AI. She is CEO of Lead with Carmen, Inc., a Global Women Peace Ambassador, and an advocate for responsible AI governance, women’s leadership, and international collaboration. Her work includes AI governance discussions at the United Nations and helping organizations navigate the opportunities and responsibilities of superintelligence.