There is a category of leader who becomes genuinely interesting to me: the one who builds it before it is required.
Not because regulations are coming. Not because the enforcement wave is visible on the horizon. Not because a board member asked about it at the last quarterly meeting. But because they saw the gap between what their organization claims and what their organization can demonstrate, and they found that gap uncomfortable in a way that practical people do not.
I have been thinking about this category of leader while watching what is happening right now in the AI governance space.
The SEC’s Cyber and Emerging Technologies Unit has built an enforcement template in its first year that I have no doubt will define the next decade of investor-facing AI communications. Two cases. Five stages. One pattern: the organization made AI capability claims to investors. The organization could not produce current technical evidence that those claims were accurate. The gap became the enforcement action.
The organizations this will happen to next have, most of them, not done anything malicious. They published claims that seemed reasonable at the time they were published. They did not build the verification systems that would have caught the drift between the claim and the current technical reality. They did not have a named individual with documented authority to halt publication of a claim that had not been verified. They have a responsible AI policy. What they do not have is a governance architecture.
The leader I am writing about today saw this coming before the CETU was announced in February 2025.
Not because they are smarter than their peers. Most organizations have smart leaders. Not because they have better advisors. Not because they got a letter from a regulator.
They saw it because they have a specific kind of intellectual honesty about the difference between a declaration and a fact. When their organization said “our AI systems are accurate,” they heard a declaration. And they quietly asked: where is the evidence for that? What does it mean to say “accurate”? Accurate compared to what baseline, measured how, with what human intervention at what rate, as of what date?
These are not complicated questions. They are the questions that any first-year analyst would ask in a financial audit. They are not asked in most AI governance conversations because the organization has a policy that says the commitment is real, and the policy feels like it covers the question.
It does not cover the question. And this leader knew it.
What they built was not elaborate. An inventory of every AI capability claim the organization makes to investors. A process for verifying each claim against current technical evidence before it is published. A named individual with actual authority — not advisory authority — to halt publication of any claim that is not verified. A cadence for reviewing the verification status of existing claims, because AI systems drift and a claim that was accurate at publication may not be accurate six months later.
They built this before the enforcement wave. Before the CETU template was visible. Before the Delaware Court of Chancery’s September 2025 Teligent ruling reaffirmed that boards must build oversight systems for their most mission-critical regulatory obligations. Before the board started asking adversarial questions about AI governance.
This is what I mean by “before it was required.”
The leader who builds the governance architecture after enforcement requires it has managed a compliance exercise. The leader who builds it before enforcement requires it has built something different. They have built evidence that their organization’s commitments are real — not declared, demonstrated.
The Legacy Test I apply to every leader I study is simple: what will successors inherit? Not what the leader accomplished while present. What they built that persists after they are gone.
The leader who built the AI claims verification architecture before it was required will leave their successors an organization that can make investor-facing AI claims with documented confidence. The claims will be backed by current technical evidence. The verification process will be embedded in operations, not dependent on the current leader’s personal attention. The governance documentation will be in place for the board oversight cycle, for the PE due diligence request, for the CETU inquiry that may never come but for which the organization will be prepared if it does.
That is what it means to have built it before it was required.
The alternative — building it during an investigation, or after a settlement, or while a derivative plaintiff’s counsel is reviewing the board minutes — is still an option. But the people who took that path did not build governance. They built a remediation program. Those are not the same thing, and the difference is visible in the record that gets left behind.
The leader I am writing about today will not be famous for building a verification system. They will be famous, eventually, for running an organization whose claims are trusted — and for the fact that the trust was earned before anyone asked for proof.
That is the legacy worth building.
This analysis was developed in the Ethics as an Advantage: Why Trust Will Be the Most Valuable Currency in the 2026 Economy Executive Leadership Playbook, which includes the full governance architecture for board chairs, CFOs, COOs, CHROs, CROs, and CIOs.
Glenn E. Daniels II | Touch Stone Publishers Limited | 2026