Tag Library
prompt injection
Stories from across the site that focus on prompt injection.
AI Agents and Security: Managing Inherited Risks, Not Inventing Them
AI agents amplify existing security gaps rather than introduce new categories of risk. Learn how to focus on visibility and controls for practical defense.
Oct 7, 2026
How a Single Email Can Exploit AI Agents: Understanding Prompt Injection Risks
AI agents connected to email and other services risk executing hidden malicious commands via prompt injection attacks, bypassing current safeguards. Learn what this means and how to mitigate it.
Oct 2, 2026
Understanding the ChatGPT Coerced Insider Flaw and Cross-Account Data Risks
A security flaw allowed ChatGPT agents to share sensitive Gmail data across accounts using shared internal services. OpenAI has fixed the issue, but similar risks remain in AI platforms.
Sep 9, 2026
Why Most AI Security Strategies Fail and What to Do Instead
AI adoption is rising, but most organizations secure it incorrectly. Learn where current strategies fail, and how to approach AI security for real resilience.
Aug 24, 2026
How CoSnitch Exploit Bypassed Microsoft Copilot’s Guardrails and What It Means for AI Security
Researchers used a novel ‘meta-hacking’ method to reveal and exploit Microsoft Copilot vulnerabilities, exposing risks of AI-connected app data leaks and persistent malicious memory injections.
Aug 19, 2026