Abstract
This disclosure describes a method of keeping accountability for AI systems and robots continuous when the person responsible leaves, whether by planned departure, sudden unavailability, incapacity, resignation or removal, and of applying the same method to the leadership roles of the organizations that operate such systems. Every accountable role recorded in a public registry, from the credentialed manager of a single system to the senior officers of an operating organization, carries an ordered chain of pre-credentialed successors. Each successor's readiness, meaning a current credential, proven reachability, reserved capacity within their span of control and a recent successful rehearsal, is verified continuously and expires if not renewed, and the registry shows publicly whether each role's succession is ready. The registry measures how much certified risk, how many systems and how many critical authorities depend on any one person, and dependence above a declared threshold must be reduced by adding successors or splitting responsibility. The outgoing person's operational knowledge, including configurations, open commitments, decision records and the credentials needed to operate, is held in a split escrow that releases to the confirmed successor only on a recorded succession event, after which the outgoing person's keys are rotated. During any gap in accountability, each affected system tightens its limits by one declared step automatically until a successor confirms. An acting successor can always stop, restore and tighten, but cannot loosen limits or make irreversible changes until fully confirmed. Removal of a person pending an investigation is recorded as a reversible interim suspension, and becomes permanent only on the investigation's findings or the person's own decision. Succession is rehearsed in advance on the systems' certified twins, and every succession event is sealed and reviewed.
Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Wang, Edward D. H., "SI Manager Succession™: Continuity of Accountability for Certified AI Systems, Robots and the Organizations that Operate Them, in which Every Accountable Role Has an Ordered Chain of Pre-Credentialed Successors Whose Readiness Is Verified and Expires, Dependence on Any One Person Is Measured and Capped, Operational Knowledge and Keys Are Held in Split Escrow, Systems Tighten Automatically During Any Gap in Accountability, Acting Successors Cannot Loosen Limits or Make Irreversible Changes, Removal Pending Investigation Is Reversible, and Succession Is Rehearsed in Advance", Technical Disclosure Commons, (November 16, 2026)
https://www.tdcommons.org/dpubs_series/13108