Keyword

prompt injection

Appears in 3 articles·Latest Oct 21, 2026

A grid of many small agent identity cards with stepped, shortening expiry tabs and risk-based access sliders, fanning out from one larger credential card, illustrating Microsoft Agent 365's agent identities and time-bound permissions.

Oct 21, 2026

How Does Microsoft Agent 365 Secure AI Agents Beyond Prompt Injection?

Microsoft Agent 365 and Project Perception give agents their own identity, time-bound permissions and risk-based access. Here is how the pieces fit.

Microsoft Agent 365Project PerceptionAI agent security

A central AI agent node with arms reaching into an inbox tray, a database stack, and an action lever, surrounded by an incomplete dashed security boundary with one open gap.

Oct 21, 2026

How to Secure Employee-Built AI Agents From Prompt Injection

Learn how to secure shadow AI agents built by employees: inventory them, limit access, add human checkpoints, and defend against prompt injection and EchoLeak-style attacks.

secure employee AI agentsshadow AI agents prompt injection

An abstract diagram of a web response packet with a hidden layer of tokens feeding directly into a sealed context module, illustrating how invisible HTTP headers get passed into an AI agent's input.

Oct 20, 2026

What Is HTTP Header Prompt Injection? How It Works and How to Stop It

HTTP header prompt injection hides instructions in response metadata that AI agents read. Learn how it works, how common it is, and a 6-step prevention checklist.

HTTP header prompt injectionprompt injection AI agent security