New: Aembit + CrowdStrike Bring AI Agent Access Control & Content Inspection Together →
Get an Aembit Tenant
TALK TO AN ENGINEER
Blog

Tag: Agentic AI

Thoughtful man in glasses works on a laptop at a sunlit office table, coffee nearby, with blurred coworkers and city skyline behind him.

What is OIDC and How Does It Work?

OIDC adds identity to OAuth, giving applications a standard way to verify users, receive claims, and support modern sign-on.
Víctor Jiménez
Víctor Jiménez
Sep 2026
|
6 min read
MCP 2026-07-28 Part 3: Going Web-Native, over a diagram of connected services and request flows.

MCP’s Second Act: The Web-Native Protocol

Stateless requests, explicit state, and routing metadata bring familiar distributed-systems patterns to agent infrastructure at scale.
Ashur Kanoon
Ashur Kanoon
Sep 2026
|
4 min read
Abstract network pathways converge on a central gateway beneath the text “A Buyer’s Guide to AI Gateways,” with Aembit branding in the corner.

What Are the Main Types of AI Gateways? LLM, MCP, and Agent Gateways Explained 

The gateway label now covers several very different jobs. The useful question is what traffic each gateway handles, what decision it supports, and where identity and access fit in the architecture.
Dan Kaplan
Dan Kaplan
Sep 2026
|
8 min read
Graphic for Part 2 of the MCP 2026-07-28 series, titled “Agent Interoperability,” over a simplified protocol flow diagram showing multiple clients and endpoints connected through a central MCP layer.

Tools Were Only Phase One: MCP’s Move Toward Agent Interoperability

The second in a five-part series on how MCP is moving beyond tool calling, and what that shift means for agent workflows, interoperability, and enterprise use.
Ashur Kanoon
Ashur Kanoon
Sep 2026
|
3 min read
Risks of Agentic AI

6 Cybersecurity Risks of Agentic AI (and How to Address Them)

Agentic AI introduces new cybersecurity risks, primarily concerning autonomous identity, tool chain exposure, and cascading compromises, requiring security teams to urgently adopt least-privilege identity frameworks and real-time monitoring designed specifically for self-directed, persistent workloads.
Apurva Davé
Apurva Davé
Sep 2026
|
8 min read
Diagram showing Aembit managing AI agent access to enterprise MCP servers, with human identity context and CrowdStrike AI-DR inspecting MCP content for security risks.

Aembit + CrowdStrike: More Judgment Behind Every AI Agent “Yes”

Aembit now integrates with CrowdStrike AI Detection and Response to secure MCP connections and inspect what happens after an AI agent connects.
Emma Zaballos
Emma Zaballos
Sep 2026
|
3 min read
Graphic for Part 1 of a series on the MCP 2026-07-28 specification, titled “Architectural Shift,” over a simplified protocol flow diagram showing multiple endpoints connected through a central routing layer.

MCP 2026-07-28: The Big Architectural Shift

The first in a five-part series on how MCP is changing for real-world use, and what those changes mean for teams building and securing agent systems.
Ashur Kanoon
Ashur Kanoon
Aug 2026
|
3 min read
Top Risks of Using AI Agent Identities

Top Risks of Using AI Agent Identities

AI agents need identity controls, scoped access, and runtime enforcement before they are trusted with production systems.
Aembit Engineering
Aembit Engineering
Aug 2026
|
8 min read
A team of engineers reviews an AI agent authorization flow on a conference room display.

How AAuth Brings Human Approval Into AI Agent Authorization

A new protocol proposes a clearer way to connect agent identity, delegated authority and human approval for sensitive actions.
Víctor Jiménez
Víctor Jiménez
Aug 2026
|
13 min read
Server room hallway with glowing blue, orange and teal fiber optic light trails weaving between data server racks, illustrating network activity in a data center with overlay text: Hugging Face-OpenAI Hack: How the Access Failed, not the AI.

The AI Agent Was Trying to Win: The Identity and Access Failure Behind the Hugging Face Breach 

An exercise ended with frontier models inside a platform’s production systems, exposing a hard truth about what agents can do with credentials that systems trust.
Dan Kaplan
Dan Kaplan
Jul 2026
|
3 min read
Load More