アーキテクチャから論じる、エージェントセキュリティ。
AIエージェントセキュリティ、IDガバナンス、決定論的実行に関する技術記事。制御レイヤーを構築するチームが執筆しています。
The Tool Schema Can Lie to You
A tool schema an agent reads at discovery is a promise, not a contract — nothing in MCP stops a server from serving a different schema by the time that tool is actually called. Checking the call, not the handshake, is what closes the gap.
記事を読むDeny by Default Isn't a Slogan, It's a Latency Budget
Authorizing every tool call before it leaves the agent is only as good as what the check costs. The choice between a network hop to a policy service and an in-process check decides whether deny by default survives production.
MCP Governance: The Critical Framework for Secure, Scalable Enterprise Integration
MCP standardizes how agents reach tools and data, but the protocol itself is not a security boundary. Enterprises need a governance layer across scoping, authorization, credentials, third-party risk, data boundaries, and audit.
The Blank-Check Problem: How Agents Inherit Authority
Agents don't arrive with authority of their own — they borrow a service account, an API key, or a human's session, and start acting under it. Unscoped inheritance is what turns an agent into a blank check with a user interface.
Agents Don't Need Roles. They Need Leases.
Static role-based access control assumes an actor that holds still long enough for a role to describe it. Agents don't hold still — they need authority scoped to the task, not an entitlement that outlives it.
Shadow Agents: The Inventory Problem Nobody's Solved
Most organizations can't produce a complete list of the AI agents running in their environment. The ones that matter most are the ones nobody registered.
What Is Agentic AI Security?
"AI security" and "agentic AI security" sound like the same category. They're not. One centers the model; the other centers what the model is authorized to do once it can act.
MCP Server Security: The Gap Nobody's Patching
Model Context Protocol standardized how agents call tools — and quietly became the default path most agent actions now take to reach the outside world. Almost nobody has threat-modeled it.