Editor's Note: Throughout the past year corporate director has partnered with Global Data Innovation to help provide accessible and practical AI governance assessments and tools to national public commissions. GDI has just been announced AI Trust Certificate Program Provide boards with documented AI governance training and frameworks so they can effectively accompany management in overseeing AI. Learn more >
Over the past year, my company, Global Data Innovation, has worked with more than 1,600 CEOs, board members, and senior leaders across 15 industries to answer a seemingly simple question: Why is Gen AI so ubiquitous yet little measurable enterprise value?
The answer was consistent. Gen AI adoption is no longer constrained by technology. It is constrained by leader preparation and governance.
Two concerns dominated the board discussion: accuracy and governance.
Why accuracy is becoming a board-level concern
Across industries, Gen AI accuracy has emerged as the biggest issue slowing or stalling adoption. Directors cited hallucinations, inconsistent output, and unpredictable performance as barriers to scaling.
Error rates for commercial generation AI tools reported this year ranged from about 26 percent to nearly 80 percent, depending on the complexity of the task.
According to research referenced in our study, 95% of executives experience negative generative AI-related incidents, resulting in billions of dollars of value being wasted or unrealized.
Boards are now asking an important question: How can you validate AI performance over time without becoming a technologist?
Why Gen AI Governance Impacts Board Credibility
The lack of a clear board-level gen AI governance framework is a pressing concern. Directors continue to ask who will oversee AI, where accountability lies, and how governance will work across silos.
Institutional investors have indicated that expectations for AI governance will increasingly be incorporated into board evaluations and voting decisions.
Gen AI governance is therefore not just a compliance exercise, but a board credibility issue.
From approval to oversight: what boards expect
Boards are becoming increasingly clear that one-time approvals are not sufficient for systems that are continually evolving.
Directors want evidence of continuous monitoring, including dashboards that track accuracy and model drift, defined escalation paths, and documented interventions or “kill switch” mechanisms.
The board doesn't want to control the model. They want assurance that accountability is clearly assigned within management and that management knows how to intervene when trust is broken.
Data integrity and leadership readiness
Any generation of AI governance discussions ultimately comes back to data. Trustworthy generative AI governance is impossible without high-quality, rights-cleared, and secure data.
Poor data doesn't just degrade models. It undermines trust.
At the same time, human readiness remains a limiting factor. Boards recognize there is a gap in AI literacy among their generation, but executives express hesitation due to competing strategic and reputational concerns.
Five governance actions for directors to 2026
1. Establish clear board-level ownership for technology oversight, including consideration of a dedicated technology or innovation committee consistent with NACD guidance.
2. Move from point-in-time approval to continuous monitoring of the accuracy and performance of the AI you generate.
3. Gen AI-related incidents require clearly defined escalation and intervention protocols.
4. Elevate data integrity, provenance, and rights management as core governance priorities.
5. Strengthen board-level GEN AI literacy to support informed fiduciary oversight.
This is not about technical mastery. It's about trustworthy, informed supervision.
For the future
Gen AI success is no longer an experiment. It's about leadership.
Boards that define accountability, insist on continuous oversight, invest in data integrity, and close governance literacy gaps will reduce risk and realize lasting ROI.
Anything that is delayed risks being dominated by events rather than design.
The age of the AI generation has arrived. Boards and CEOs now have an opportunity to step up and lead.
