How the Network Governs Its Agents

The group grades decisions by consequence rather than by capability, and every agent has a named human owner who answers for its output.

Governance is what makes an agent workforce ownable

A network with agents acting inside it has to answer three questions on demand: who authorised this, what were they allowed to do, and what happened. A group that can answer them owns its agent workforce. One that cannot merely hosts it.

Across ten properties this stops being philosophical quite quickly. An agent moving value on Flashy Finance and an agent reading a completion event from Flashy Academy are governed by the same rulebook, or they are governed by nothing that survives contact with a regulator.

The consumer network's particular exposure

Enterprises discover governance gaps in an audit. Consumer networks discover them in public, at speed, from people who screenshot. That difference sets the bar: the record has to be good enough to answer an individual citizen, not just an annual review.

It is also why the network wrote its rules down before it needed them. The Flashy Constitution is the standing text; Why an AI Workforce Needs a Constitution is the argument for having one at all.

What a citizen is owed

One thing, and it is more demanding than it sounds: if an agent affected your balance, you are entitled to know that an agent did it, under whose authority, and how to contest it. A network that cannot say that is asking its population to trust software it will not name.

Where this is documented

The operator account — how this behaves when you are running the fleet — lives with FlashyOS, the agent mesh the group's organizations run on. The institutional definition and what it is worth lives with GDA Group.

One claim, one canonical home: this page states how the network uses it and links the rest.