The Linux Foundation’s Agentic AI Foundation is bringing developers and industry leaders together in Tokyo to address the standards, security and infrastructure needed for production-ready AI agents.
The Agentic AI Foundation (AAIF), hosted by the Linux Foundation, will hold AGNTCon + MCPCon Japan 2026 on September 10–11 in Tokyo. The two-day event will bring together developers, researchers and organisations working on agentic AI to examine how AI agents can move beyond experimental projects into reliable production systems.
The conference will focus on technologies and infrastructure required to build interoperable agent environments. More than 40 sessions will cover areas including multi-agent coordination, tool integration, sandboxing, evaluation, observability, security and enterprise governance. These discussions will address practical challenges such as defining trust boundaries when agents interact with each other and managing approval processes and audit trails.
A major focus will be the use of open-source technologies for building agent infrastructure. Sessions will explore how components such as orchestration layers, identity and authentication systems, and observability tools can be combined to create production-ready environments. The event will also examine the Model Context Protocol (MCP) and its role in connecting AI agents with external tools and services.
Dedicated tracks at AGNTCon and MCPCon Japan will cover dependable agent systems, testing and evaluation, enterprise adoption, human-agent collaboration and distributed multi-agent architectures. These sessions will examine approaches for improving reliability, implementing guardrails and sandboxed environments, monitoring agent behaviour and evaluating performance before deployment.
AGNTCon and MCPCon Japan form part of the broader efforts of the Agentic AI Foundation to develop vendor-neutral infrastructure and standards for agentic applications. By bringing developers and organisations together around shared technologies and protocols, the conference aims to encourage interoperable solutions that can operate safely across different applications, data sources and platforms in production environments.
