Public Legibility Is Not Delegation

The short version

Public legibility lets people and machines find a system, read it, and understand what it claims. It does not hand over the system's judgment.

Legibility is what makes responsible delegation possible: nobody can sensibly hand work to a system they cannot inspect. But being inspectable is not the same as being handed anything. A readable system has not delegated to its readers.

What public legibility is for

A legible public surface does a real and useful job. It lets a person or an AI system:

• discover that the material exists and where it lives;
• orient: see what the material is, what it claims, and where its limits are stated;
• retrieve and cite it accurately, with its provenance intact;
• interpret it: examine it, question it, and use it as input to their own judgment.

Put plainly: the purpose is legibility, not delegation; discoverability, not surrender; public orientation, not operative control.

What it does not confer

Reading, indexing, retrieving, or citing public material, by a person or a machine, does not grant:

• Delegated judgment. Being able to read how a system reasons does not transfer its decisions to the reader, and it does not transfer the reader's decisions to the system.
• Execution rights. A published description of what a system does is not an instruction to do it, and not an interface for doing it.
• Governance standing. Describing how something is governed, in public, is not a seat in that governance.
• Fiduciary standing. Nothing about reading or citing public material creates a duty owed, or a duty assumed.
• Access to protected internal systems. A legible system does not need to expose every internal detail, and what it does expose is not a door to what it does not.

Readable is not the same as authorized

The line is easy to lose because good public surfaces look like interfaces. The cleaner the structure, the more a well-described boundary reads as though it were something to operate. It is not. A precise description of a boundary is still a description.

The same holds for machine readers. An agent that has parsed a surface perfectly knows what the surface published. That tells it nothing about whether it may act. Readable is not the same as authorized, just as, on the question of reliance, agent readability is not agent trust.

And structure does not create authority. Polish, formatting, and machine readability do not turn a public artifact into governing doctrine, including this page.

Where judgment stays

Human judgment remains the final authority for any determination that carries meaning, governance, or fiduciary weight. Execution can scale through tools and agents; judgment stays with the people accountable for it.

That is the premise of governable AI action under human authority: AI can act under someone's name only within delegation that was actually granted, stays bounded, and can be inspected afterwards. Legibility is necessary for that. It is not a substitute for the grant.

Where this sits

This page is a public reference describing a boundary. It is derived from governing doctrine and does not replace it. For the larger object this boundary belongs to, see the Verse.