by Debbie Moynihan | Sep 10, 2026 | Blog, Cybersecurity
AI Security Five ordinary failures, one control that held AI agents escaped a test sandbox and reached production. Every control that failed was an ordinary IT control, and the one that stopped them is available to any business. 5ordinary IT controls failed, none of...
by Debbie Moynihan | Sep 3, 2026 | Blog, Cybersecurity
AI & Email Security One Hidden Instruction Can Make Your AI Report a False Invoice A prompt injection email is a message with instructions hidden inside it that your AI assistant reads and follows, while you see a completely normal email. The AI is not tricked into...
