From AI use to AI authority: introducing OBO Governance
- Neil Hare-Brown
- 3 days ago
- 1 min read
…and breathe.
The governance challenge created by AI may feel entirely new. But it may not require an entirely new solution.
As AI moves beyond productivity support to negotiate, authorise, communicate and execute business processes, the central issue is no longer simply intelligence or technology. It is delegated organisational authority.
AI does not merely democratise knowledge; it democratises capability. Every employee can now access sophisticated, expert-level assistance—and potentially act with far greater consequence on behalf of their organisation.
My paper proposes an On-Behalf-Of—or OBO—Governance model.
When AI materially represents an organisation or exercises organisational authority, it enters the OBO State:
The organisation remains the principal.
An authenticated individual invokes and supervises the authority.
The AI operates as a computational proxy within a defined Authority Profile.
Authority is delegated, never transferred.
Accountability remains with the human and organisational principals.
This approach builds upon principles that have governed human society for centuries: agency, proxy, delegated authority, revocation and accountability.
OBO Governance does not replace ISO 42001, the NIST AI RMF or existing AI regulation. It adds a missing layer: Authority Governance.
The question is no longer merely, “How should employees use AI responsibly?”
It is: “How should organisations govern thousands of AI-assisted acts undertaken on their behalf every day?”
Perhaps the foundations of effective AI governance already exist.
Download the full paper
Explore the OBO Governance model in full.
Includes how AI can be governed as a computational proxy operating under clearly defined delegated authority, accountability and organisational control.