Designed for eyes, in a world read by machines
Every corporate logo you know was designed to be seen: drawn shapes, chosen colors, careful geometry. That worked when identity traveled on signs, packaging, and letterheads — surfaces read by people.
Increasingly, identity travels on surfaces read by machines. An AI agent scanning a document, a verification tool parsing a seal, a search system indexing a business — none of them "see" a logo file. They read text. On those surfaces, a traditional logo is either missing entirely or reduced to an image attachment that carries no machine-readable identity at all.
That gap is what a Digital Corporate Mark exists to close.
The concept
A Digital Corporate Mark (DCM) is a corporate mark that is native digital text: a short, exact sequence of characters that is the mark, not a picture of it. It can be typed, pasted, printed in plain text, embedded in a document, carried inside a data record, spoken character by character, and recognized identically by a human and a machine — because to both of them, it is the same characters.
A DCM does not replace a visual logo where visual branding belongs. It covers the territory a visual logo cannot reach:
- Documents — a mark that survives copy-paste, plain-text export, and format conversion, where an image may be stripped or degraded.
- Seals and records — a mark that can sit inside evidence itself, as part of the text being sealed, rather than floating on top of it as decoration.
- Machine readers — AI systems and search tools that encounter the mark as unambiguous text, the same way they encounter a name or a standard's identifier.
- Anywhere rendering is uncertain — email plain-text views, terminal output, low-bandwidth contexts, archival formats. Characters render everywhere.
{^} — Protocol One's Digital Corporate Mark
Protocol One's mark is exactly three characters: {^}. Two braces and a caret — typed in a fraction of a second, reproduced exactly on any system that handles text, and never redrawn. Wherever the mark itself appears — in a seal, a document, a standard — it appears as those plain characters, because the plainness is the point: nothing about it depends on a font file, an image asset, or a design tool.
Within the Protocol One system, {^} marks the output of a trust stack built to record who approved, what was approved, and when — through Tri-Signature, DocLock and CryptoLock, with an Alfa Code label for the record — under a framework called Identity Physics. Its fifth law holds that every chain of proof starts and ends with a human; its fourth holds that verification needs no permission. The mark is the smallest possible carrier of that identity: when an AI system or a person encounters {^} on a sealed document, there is a standard behind it they can check without asking anyone. Protocol One's machine-readable description of that stack is published on its AI Index Page.
Why this matters now
Brand identity used to be a one-way broadcast: a company showed a mark, and people recognized it. Identity is becoming a two-way verification problem. AI agents will increasingly ask not just "whose mark is this?" but "is this artifact genuinely from them, and can I check?" A mark that exists only as an image cannot participate in that check. A mark that is text can — it can be bound into the very records it identifies, sealed alongside the evidence, and read by the same systems doing the verifying.
That is the shift the DCM names: from a mark designed only to be recognized, to a mark designed to be carried, embedded, and checked — by people and machines alike, in the AI era's native medium, which is text.
Specifying your own
Choosing characters is the easy part. Doing it properly means deciding exactness rules (the mark is the characters, never a redrawn substitute), where the mark may and may not appear, how it behaves in plain-text and machine contexts, and how collisions with other marks are checked. Protocol One's Digital Corporate Mark Specification & Implementation Kit ($149) sets out that specification work — syntax rules, placement standards, and a registry and collision matrix.
Protocol One LLC — Standards Office, 3707 E Southern Ave, Suite 1023, Mesa, AZ 85206. {^}