100:00:00,000 –> 00:00:02,860A team asked an AI agent to just wire up Microsoft Graph 200:00:02,860 –> 00:00:05,880in 10 minutes later they had a working app and directory, 300:00:05,880 –> 00:00:07,780read
100:00:00,000 –> 00:00:02,860A team asked an AI agent to just wire up Microsoft Graph 200:00:02,860 –> 00:00:05,880in 10 minutes later they had a working app and directory, 300:00:05,880 –> 00:00:07,780read
Summary Imagine your security operations center (SOC) waking up to an alert: “Copilot accessed a confidential file.” It’s not a phishing email, not malware, not a brute force attack —
🔍 What This Episode Covers In this episode, we explore: Why AI agents behave unpredictably in real production environments The hidden risks of connecting LLMs directly to enterprise APIs How
Summary You know that sinking feeling when your Power Pages form refuses to validate and the error messages feel basically useless? That’s exactly the pain we’re tackling in this episode.
100:00:00,000 –> 00:00:03,180Most organizations come to fabric with the same comforting assumption. 200:00:03,180 –> 00:00:05,760If we get the permissions right, the numbers will be right. 300:00:05,760 –> 00:00:08,760They are wrong,