Google and Microsoft Jointly Launch ARD Protocol; Anthropic and OpenAI Absent
Alina Collins
Google, Microsoft and other legacy tech giants jointly launched the ARD agent-discovery protocol. Anthropic and OpenAI are both absent — not over a technical disagreement, but over who gets to be the front door to enterprise AI.
What problem does ARD actually solve?
ARD — Agentic Resource Discovery, a unified standard that lets AI agents automatically find and call software capabilities — aims to turn legacy enterprise platforms into unified AI gateways.
In plain terms = an employee kicks off an AI task inside GitHub Copilot, Google Gemini or Salesforce CRM, and the system automatically locates and invokes every ARD-compatible AI service, with no manual tool-switching.
Microsoft stressed that ARD handles "discovery" only — not authentication, authorization or governance. This means → it is a directory layer, not a set of keys.
How does it relate to Anthropic's MCP?
MCP — Model Context Protocol, a standard Anthropic released last year that lets AI agents access external app data, functioning much like an API — already has support from Microsoft, OpenAI and Google.
ARD is seen as an extension of MCP's vision: MCP solves "how AI connects to outside data"; ARD solves "how AI discovers which capabilities exist to connect to."
This means → ARD does not replace MCP. It adds an automatic discovery layer on top — upgrading from "can connect" to "can find what to connect to."
Why are Anthropic and OpenAI sitting this out?
Both companies are building Claude and ChatGPT to be the primary gateway through which employees reach all AI tools and enterprise apps — they want to be the center of the "superagent" (a general-purpose AI system that executes tasks across applications).
Joining ARD would mean accepting that Microsoft's or Google's platform is the real gateway, reducing themselves to a component inside someone else's ecosystem — a direct conflict with their strategy.
This reflects the deepest fault line in enterprise AI: legacy software giants want to integrate AI around themselves, while AI-native companies want to integrate software around themselves — two competing "center of gravity" models that cannot coexist.
Where does this gateway battle stand now?
ARD has just launched; whether it wins broad adoption remains highly uncertain — the protocol's technical merit does not guarantee market acceptance.
This means → the current lineup (legacy giants vs. AI-native firms) is more of a strategic statement than a settled outcome.
Put simply = the fight over who owns the front door is still at the flag-planting stage — the real tipping point is which side makes enterprise customers unable to leave its ecosystem first.
Content is for reference only, not financial advice.