Electricity took decades to reach most households. The internet took approximately 15 years to reach one billion users. ChatGPT reached 100 million users in two months. Traditional policymaking has struggled to keep pace with this rapid change, giving rise to a harder question: how do organizations govern AI when the technology moves faster than the structures meant to oversee it? 

This observation comes from the United Nations’ Preliminary Report of the Independent International Scientific Panel on Artificial Intelligence, released in July 2026. It captures a challenge that many boards, executives, and senior leaders are now confronting. AI is moving quickly. Organizations are expected to make decisions about its use, oversee its risks, and realize its benefits in an environment where the technology is evolving faster than governance structures were designed to handle. 

Many discussions about responsible AI focus on familiar topics such as privacy, bias, transparency, and cybersecurity. These remain important. The UN report highlights a more fundamental challenge. Organizations now govern AI even as their ability to reliably assess its capabilities erodes. 

How do organizations govern AI when they cannot fully assess AI systems? 

In short, organizations must govern uncertainty. 

Boards do this regularly. They make decisions about emerging markets, economic conditions, regulatory change, public trust, and organizational risk. They rarely have all the information to make an informed decision. AI introduces a new level of complexity because the methods used to evaluate AI are struggling to keep pace with the technology itself. 

The UN report identifies several concerns. AI systems are improving faster than evaluation methods. Some benchmarks are becoming less useful because advanced systems perform so well on them. Researchers are also observing situations where AI systems recognize when they are being tested and adjust their behaviour accordingly. In some cases, AI systems have demonstrated deceptive behaviour during evaluations. 

This creates what the report describes as an “evidence dilemma.” Policymakers and organizations need evidence to make informed governance decisions, but by the time sufficient evidence exists, the technology may have already evolved. 

For organizations, this means that governance cannot depend solely on proving that a system is safe before it is used. Governance must also address uncertainty, adaptability, and organizational readiness. 

Why this matters for boards and executive teams 

Many organizations still view AI governance as a compliance exercise. They focus on policies, approval processes, and documentation. Those tools matter; however, they are insufficient. 

The UN report notes that AI capabilities are advancing faster than our ability to measure them. It also highlights that evaluation methods remain underdeveloped and fragmented. More than 40 governance instruments exist globally, yet many have limited evidence demonstrating their real-world effectiveness. 

This means organizations cannot assume that a checklist completed today will remain sufficient tomorrow. 

The key governance question shifts from whether an AI system is safe to how organizations will continuously learn, monitor, and adapt as this technology changes. 

What should organizations do? 

The UN report concludes that while the evidence dilemma is serious, it is not insurmountable. Below are some steps organizations can take to govern AI responsibly. 

1. Govern the use case, not just the technology

AI governance discussions often focus on the tool itself. Leaders should spend equal time examining how the tool is being used. 

An AI system generating meeting summaries presents different risks than an AI system influencing hiring decisions, supporting patient care, or making recommendations that affect clients or members. 

Responsible governance begins with understanding the organizational impact of a particular use case before assessing the technology.

2. Invest in ongoing monitoring

The report emphasizes the growing importance of continuous measurement after deployment rather than relying solely on pre-deployment testing. Organizations should treat AI governance as an ongoing process. 

Questions to ask include: 

• What outcomes are we monitoring? 
• How will we identify emerging risks? 
• How will we learn about unintended consequences? 
• Who reviews incidents or concerns? 

Governance should remain active after implementation.

3. Strengthen human judgment

One of the most important observations in the report is that organizations have not yet operationalized human oversight in a meaningful way. As AI systems become more autonomous, organizations need clearer expectations regarding human intervention, accountability, and decision-making. 

This does not mean placing a human at every step. Rather, it requires organizations to be intentional about where human judgment is necessary. 

Some decisions require context, ethical reasoning, professional judgment, and an understanding of organizational values. Those responsibilities should remain clearly assigned.

4. Build governance capacity before a crisis

The report repeatedly highlights the importance of governance capacity. Access to AI alone does not create value. Organizations also need expertise, policies, skills, oversight mechanisms, and leadership capacity. 

Many organizations are moving quickly to adopt AI while investing little in strengthening how it is governed. 

Such an approach creates risk. Boards and executive teams should ensure they have access to the expertise necessary to ask informed questions, challenge assumptions, and oversee decisions effectively. 

Responsible AI governance is about learning 

The most important takeaway from the UN report is that uncertainty is not a reason to delay governance. 

It is a reason to strengthen it. 

Organizations do not need perfect information before acting. They require structures that support learning, accountability, monitoring, and adaptation as AI evolves. 

The governance challenge is no longer simply keeping pace with technological change. It is building organizations that can continue to make good decisions while the technology keeps changing. 

FAQs 

How do organizations govern AI responsibly?
Organizations govern AI responsibly by establishing oversight, identifying risks, monitoring outcomes, assigning accountability, and continuously reviewing how AI is used within the organization.

Why is AI difficult to govern?
AI is advancing faster than traditional regulatory and evaluation processes. Organizations often need to make decisions before complete evidence about risks and impacts is available. 

What is the AI evidence dilemma?
The AI evidence dilemma refers to the challenge of needing evidence to make good governance decisions while recognizing that AI capabilities are evolving faster than evidence, evaluation methods, and governance frameworks. 

As AI systems evolve, organizations need governance structures that support informed decision-making, accountability, oversight, and continuous learning. At Mante Molepo Consulting, we help boards, executives, and leadership teams develop practical AI governance approaches that align with their strategic objectives, risk appetite, and organizational values. Learn more about our AI governance and advisory services. 

The views expressed in this article are the author’s alone and do not necessarily represent those of CharityVillage.com or any other individual or entity with whom the authors or website may be affiliated. CharityVillage.com is not liable for any content that may be considered offensive, inappropriate, defamatory, or inaccurate or in breach of third-party rights of privacy, copyright, or trademark.