The Agentic AI Foundation is establishing neutral, open standards to enable enterprises to deploy artificial intelligence agents at scale. The initiative brings together technical frameworks including MCP, A2A, and Goose—projects designed to allow agents, tools, and systems to work seamlessly together across organizations.
Angie Jones, the foundation's VP, draws from extensive experience leading AI adoption at Block, where she spearheaded efforts to teach AI agent usage across the entire 12,000-person organization. Unlike traditional tech rollouts targeting engineering teams, Block's approach required training finance, marketing, design, and HR teams on how to leverage agents in their workflows. The company developed its own internal agent, Goose, to automate engineering tests and coding tasks, but realized that moving beyond initial adoption to building genuinely autonomous operations required deeper organizational change.
The foundation's work addresses a critical industry gap: while many enterprises are beginning to experiment with AI agents, few have established the technical standards or organizational practices needed for large-scale deployment. By promoting neutral standards and sharing lessons learned, the foundation aims to help organizations balance AI autonomy with human oversight while avoiding vendor lock-in.
Key Points
Agentic AI Foundation establishes open standards (MCP, A2A, Goose) to enable interoperable AI agent deployment across enterprises
Organization-wide AI adoption requires training all business functions, not just technical teams, on AI agent capabilities
Enterprises must progress beyond initial AI fluency to building autonomous systems that drive measurable productivity improvements
Neutral standards prevent vendor fragmentation and help organizations make strategic choices about AI agent deployment
Balancing AI autonomy with human oversight remains a key challenge as enterprises scale agent adoption