The Committee Fallacy

Each section moves in three tiers: a board line, a why it matters note, and a shaded detail block carrying the authority.

The committee is the governing body, not the governance

This paper names the error, documents how quickly institutions are committing it, specifies the anatomy of the program the committee must actually govern, and answers the speed objection directly. The layering is precise because the precision is what makes governance defensible: the committee governs the program, the program guides the AI activity, and the record proves both.

The fallacy, named and documented

1.1 Why committee designation feels like governance

Assigning a risk to a committee produces a documentable act and signals attention, but neither is governance. Governance is a continuing activity; the charter amendment is its precondition, not its performance.

Why it matters. The distinction would be pedantic if Delaware had not built two cases on it. Clovis is the controlling instruction: a board with a committee overseeing its lead clinical trial was liable because it knew of reporting problems and did nothing. Boeing is the threshold half: no charter named the risk, and silence read as oversight absence. Only the charter that is followed, in documented committee activity, protects.

1.2 The gap, measured

Institutions are naming and assigning AI risk far faster than they are building the program beneath it. The charter amendment without a program moves the institution from the first prong of the doctrine to the second.

Why it matters. The first prong, no system at all, is becoming unavailable to plaintiffs. The second prong, a system not monitored, is where the AI cases will be fought, and the disclosure data shows institutions walking straight into it.

The charter amendment moves the institution from Boeing's exposure to Clovis's. The first prong asks whether the structure exists. The second asks whether it governs. Structure is the easy prong.

The anatomy of the program

2.1 What the committee governs

The committee governs the program: a defined, documented system establishing what must be governed and who is accountable, leaving how to management, counsel, and operators. Crossing into the how dissolves the committee's position above the activity.

Why it matters. The Center's published architecture gives the anatomy a name and structure: five pillars, fourteen domains. The pillars and domains are public; the measurement layer beneath them is the work of the Center's assessment instruments. The anatomy alone is sufficient here, because the anatomy is what the case law describes from the outside.

2.2 The five pillars and fourteen domains

2.3 The anatomy against the case law

The architecture answers Boeing element by element. The committee with charter responsibility is Domain 2; the reporting cadence and protocols are Domains 2 and 13; the independent channel is Domain 13's design requirement; the documented red-flag response is Domain 13 in execution and Domain 14 in verification.

Why it matters. The case law describes the program's absence; the program is the case law's specification, run in reverse and in advance. The overarching question the program teaches its committee to ask is not whether a specific model is safe, but whether the program is designed and operating effectively so the institution can make informed decisions on risk and reward. That is the Informed Decision Standard in interrogative form.

The objection: will the program slow us down

The objection that fourteen domains means bureaucracy is backwards. Ambiguous obligation is the brake on AI adoption; clear obligation is what lets AI move.

Why it matters. Where no one knows who can approve a deployment, what standard it must meet, or what happens when it misbehaves, every decision escalates to everyone and velocity collapses. The deployment that meets the Pillar III standards, inside the Domain 4 appetite, with the Domain 13 triggers attached, can be approved by the accountable owner without convening the institution. The program is the reason the institution can say yes quickly and defend the yes later.

The program is the governance

The committee that governs the program is doing what Delaware, the examiners, and the investors are all separately asking to see. The committee that merely exists is the fallacy with a charter.

The anatomy is not mysterious and never was: a mandate, a yardstick, standards, strategic alignment, and a voice; five pillars, fourteen domains; what is governed and who is accountable, with how left to the layers built for it. The doctrine has been explicit about which one it tests, and the disclosure data shows the test approaching at scale. The program is the governance.