Three pieces this week. Each one separates a lookalike for governance from the record an examiner can use.
A claim measured on a private corpus. A detector that reports drift. A plan that pre-authorizes a change. Each one resembles governance. The record is the part that survives the question.
This week.
If the Corpus Is Private (June 8). Vendor demos cite numbers measured on telemetry no counterparty can see. The corpus is internal. The pipeline is opaque. The trace is gone. A number measured on private data cannot fail in public, which is the same as saying it cannot be trusted in public. The discipline that makes a claim usable to a counterparty is old and unglamorous. Stand the pipeline on a corpus the counterparty can download, and let the number survive or not.
Drift Detection Is Not Governance (June 10). Drift detection is shipping across the agentic-AI stack. It reports that behavior moved. It does not decide what the agent was permitted to do once it moved, and it does not leave a record an examiner can read. A detector is a sensor. A governor sits in the decision path, sets the authority of the action before it executes, and signs what it decided. The examiner does not ask whether you noticed the drift. The examiner asks what you did and where the record is.
The Admission Gate (June 12). Most governance watches the action and reads it against the policy in force. A second moment goes unwatched. The moment the model is permitted to change. A learning event is not a decision. It is the model rewriting the function that produces decisions, and it resets the disposition behind every decision that follows. The FDA's Predetermined Change Control Plan names the change and pre-authorizes an envelope. A plan is not a record. The runtime evidence of which change committed, against which bound, is the artifact almost no one is producing.
What changed.
Federal banking examiners moved AI into the exam itself. Reporting this month describes AI-governance questions entering standard bank examinations, while SR 26-2 still carves generative and agentic AI formally out of scope. Industry surveys this spring put a majority of banks unable to confirm they could halt a malfunctioning AI model or report the failure to a regulator. The examiner is asking the question the carve-out called premature. The distance between deploying the model and proving control over it is now an examination item.
The EU Digital Omnibus is moving from agreement to law. The high-risk obligations defer toward December 2, 2027 for stand-alone Annex III systems and August 2, 2028 for AI embedded in regulated products, but only on formal adoption and publication in the Official Journal, expected before August 2, 2026. The Article 50 transparency obligations stay on the original schedule. The high-risk runway lengthened. The transparency date did not move.
The NAIC AI Systems Evaluation Tool pilot runs March through September across twelve states. The tool is expected to be adopted at the Fall National Meeting on pilot feedback. The state insurance examiner is still on track to move before the federal banking examiner finishes writing its own rule.
What we are tracking.
The runtime-authorization wave. A cluster of agent-governance launches now enforce policy and authorization at the tool call. Enforcement answers whether the agent may act. It does not produce the signed, replayable record of what the agent did, and it does not gate the moment the model is permitted to change. Enforcement without a record is half the control loop.
The interagency RFI. The banking agencies have signaled a request for information that addresses AI, generative AI, and agentic AI directly. The carve-out points institutions back at their own risk practices until it lands. Where that RFI draws the line between model risk and agent assurance is the line we are watching.
The Official Journal date. Provisional agreement is not law. The revised high-risk timeline becomes real on publication, and until then the deadlines on the books are the old ones. The August 2 transparency obligation is the near date that did not move.
Thirty minutes. Architecture, not sales.
A conversation about what the trajectory record has to contain to survive the next examination cycle, and where the artifact should live in your organization.
JonathanLuethke@WayfinderSystemsGroup.com
