Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

In this episode of Data Security Decoded, host Caleb Tolin is joined by Camille Stewart-Gloster, CEO of CAS Strategies and former Deputy National Cyber Director, to unpack how AI is redefining cyber risk at every layer of the organization. Camille explains why identity-based attacks are so effective and how non-human identities (from APIs to AI agents) are quietly expanding the attack surface. She emphasized how critical MFA is for organizations to enable as they scale up AI operations, and why conditional access and governance must be foundational, not optional.

Stop building AI agents that CISOs will never approve. If your agents are stuck in the POC (Proof of Concept) stage, it’s likely because they lack a "Passport" and a governance framework. In this clip, Arjun Subedi breaks down why "how well it works" isn't the biggest question in AI anymore—it's "how can I govern it?" Discover how mapping AGENTIC attacks to the MITRE ATT&CK framework through SafeMCP is the missing link to enterprise-level deployment.

AI is everywhere—writing emails, creating videos, even cloning voices. But artificial intelligence also comes with real risks, including privacy concerns, deepfakes, and smarter online scams. Artificial intelligence learns by spotting patterns in massive amounts of data—and that power can be misused. AI tools may collect personal information, create realistic fake content, or help scammers craft messages that look completely legit.

Can you trust the code AI generates? In this video, we build a custom AI Security Benchmarking tool to put models like Gemini, Mistral, and GLM 4.5 to the test. Using Windsurf, OpenRouter, and Snyk, we automate a pipeline that prompts multiple LLMs to write an application, then immediately scans the output for security vulnerabilities.

Celso explains how Markdown for Agents was conceived, built, and shipped in just one week, why AI systems prefer markdown over HTML, and how converting a typical blog post from 16,000 HTML tokens to roughly 3,000 markdown tokens can reduce cost, improve speed, and increase accuracy for AI models. We also explore Moltworker, a proof-of-concept showing how a personal AI agent originally designed to run on a Mac Mini can instead run on Cloudflare’s global network using Workers, R2, Browser Rendering, AI Gateway, and Zero Trust.