With AI agents moving from research projects to production systems, achieving interoperability across tools and platforms has become critical. Angie Jones, VP of the Agentic AI Foundation, joined the Practical AI podcast to discuss how open standards like MCP, A2A, and Goose are laying the groundwork for a cohesive agentic ecosystem. The foundation's work centers on creating neutral, vendor-agnostic standards that enable agents, tools, and systems to communicate seamlessly at enterprise scale.
Beyond technical standards, the conversation explored the organizational challenges of agentic AI adoption. Jones outlined what it takes to drive AI agent implementation across large organizations, emphasizing the importance of neutral standards that prevent vendor lock-in and foster genuine interoperability. The episode also examined how global markets are approaching agentic AI differently, reflecting varied regulatory environments and business needs across regions.
A key theme emerged around human-AI collaboration: as organizations increasingly delegate work to AI agents, finding the right balance between autonomy and human oversight becomes critical. The discussion underscored that technical standards alone aren't sufficient—organizations must also establish governance frameworks and cultural practices that enable teams to effectively manage and oversee autonomous AI systems in production environments.
Key Points
The Agentic AI Foundation is developing open standards (MCP, A2A, Goose) to enable interoperability between AI agents, tools, and systems at enterprise scale
Organizational adoption of agentic AI requires both technical standards and cultural/governance changes to balance human oversight with agent autonomy
Neutral, vendor-agnostic standards are essential to prevent lock-in and enable genuine ecosystem collaboration across the industry
Global perspectives on agentic AI vary significantly, shaped by different regulatory environments and business requirements across regions