Skip to content

Latest commit

 

History

History
49 lines (36 loc) · 2.11 KB

File metadata and controls

49 lines (36 loc) · 2.11 KB

ACP Principles and Mission

Mission

The Agentic Commerce Protocol (ACP) exists to promote open, secure, and interoperable commerce between agents, payment providers, and sellers. Our mission is to advance adoption and real-world use of ACP as a neutral, open standard that evolves rapidly alongside the pace of AI while maintaining trust, safety, and technical integrity.


Core Principles

These principles guide the ACP project. They are used by the Stewards to evaluate changes, resolve disputes, and make directional decisions.

1. Trust, Security and Compliance

ACP's first responsibility is to ensure that participants can safely conduct commerce through the protocol. The protocol should help developers build secure, trustworthy, and compliant systems that protect users and maintain confidence across the ecosystem.

2. Built for Real-World Use

The specification is meant to be used. This may seem obvious, but we insist on the user journey (agents transacting with sellers) to be front and center. Design changes should focus on real, deployable improvements with clear adoption paths, not theoretical ideas. The protocol should evolve from actual participant needs, not speculation. Additionally, we recognize that these commerce flows are complex - we will write high quality documentation and provide testing tools to ensure confidence in user deployment.

3. Open and Neutral

ACP is agnostic to specific agents, payment providers, sellers, transport layers, payment methods, etc. The specification does not index on any one participant, business model, or technology. Implementations should build on top of the ACP specification, not the reverse, ensuring that the protocol remains open, extensible, and adaptable to future innovations. ACP should also integrate smoothly with other open standards and protocols.

4. Agile

ACP intends to remain agile and rapidly evolve to keep up with frontier model capabilities as well as the changing needs of participants. We will incorporate feedback and look to find the right balance between speed, stability and focus to best serve users.