Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

Day 2 at the AI Impact Summit was all about debunking myths. One major takeaway from our conversations today: Most leaders think AI security is just about stopping 'bad prompts.' But the real danger is exposing sensitive data to the model in the first place. If you aren’t sanitizing your data before it hits the AI, you’re leaving the door wide open. We’ve been showing attendees at Bharat Mandapam how Protecto bridges the gap between basic AI security and true Data Privacy.

This video shows how Claude Code handles a full tenant setup in a single workflow: creating a new organization, deploying Sigma rules, and enabling Git Sync, all in one run with LimaCharlie. Claude Code creates the organization, deploys the community Sigma rules available on the platform, and configures the Git Sync extension. That extension automatically creates a GitHub repository and syncs both the detection rules and the full tenant configuration to it, giving you version-controlled infrastructure from day one.

Shadow AI models with weaker guardrails increase the power of already talented attackers and help less skilled users assemble dangerous campaigns. Frontier systems write malware when prompted as penetration tests, scale up attack speed and make it easier than ever to launch offensive operations against defenders. ⸻ For more information about us or if you have any questions you would like us to discuss email podcast@razorthorn.com. We give our clients a personalised, integrated approach to information security, driven by our belief in quality and discretion..

Before you hit “send” on Zelle, Venmo, or CashApp — pause. Most peer-to-peer payment apps work like digital cash. Once the money is sent, it’s often gone for good. No undo button. No guaranteed fraud protection. Don’t let a quick payment turn into a permanent loss.

Your code is one of your organization’s most critical assets, and protecting it should be as resilient as the applications it builds. In this video, we walk through how Rubrik delivers enterprise-grade protection for enterprise-critical code by securing Azure DevOps repositories against accidental deletion, malicious activity, and ransomware threats.

See how CrowdStrike Falcon Fusion SOAR Test Mode lets security teams validate workflows instantly—without live alerts or production risk. You’ll learn how to run playbooks end-to-end using past events or mock data, safely test high-cost actions like LLM analysis, and debug logic with full visibility. With built-in transform testing and mock inputs, Test Mode speeds up automation and reduces errors for faster, safer deployment.

See how CrowdStrike Falcon Next-Gen SIEM enables security teams to deploy log collectors instantly via the Falcon Sensor—without relying on traditional software distribution processes. In this demo, you’ll learn how to create a Log Collector Policy, automatically install collectors across host groups, validate installation through Falcon telemetry, and centrally manage third-party data ingestion from Fleet Management.

Your AI tools just became the perfect hiding spot for hackers. Cybercriminals have found a new attack vector: weaponizing the AI assistants your team uses every day. In this live demonstration, we show how adversaries can turn tools like Claude into persistent backdoors and how CrowdStrike Falcon stops them cold. TIMESTAMPS: WHAT YOU'LL LEARN:– How hackers exploit AI automation features to create backdoors– Why password resets and patches won't stop this attack– How behavioral detection catches threats hiding in legitimate tools– Real-time threat prevention in action.

Ever used incognito mode and thought, “Nice, now no one can see what I’m doing?” The truth is, incognito mode doesn’t make you invisible online. Incognito mode helps keep your activity private on your device by not saving your browsing history, cookies, or autofill data once you close the window. That’s great for shared devices or logging into multiple accounts—but it doesn’t stop tracking beyond your browser.

Agentic AI is fundamentally changing how software operates, and in doing so, it breaks the identity and access models that many organizations still rely on. Unlike traditional applications, agentic systems are non-deterministic, long-running, and capable of autonomous decision-making across infrastructure, data, and production services. These systems do not fit within legacy identity assumptions built for humans, static workloads, perimeter controls, or long-lived credentials.