Just came across an interesting report from Hudson Rock. A threat actor claims to have extracted millions of enterprise records from Azure tenants. Some of the organizations reportedly affected: - McDonald’s: ~1.7M+ records - TCS: ~800K+ - Vodafone: ~425K+ - HCL Technologies: ~250K+ - IHG: ~185K+ - Kyndryl: ~170K+ - Gap: ~80K+ - Hexaware: ~20K+ - Wyndham Hotels: ~9K+ The exact attack vector is not yet confirmed, but compromised credentials, stolen session tokens, phishing or overly permissive integrations are among the possibilities. Another reminder of how important identity security is in cloud environments. Worth a read 👇 https://www.hudsonrock.com/blog/massive-azure-exfiltration-campaign-exposes-millions-of-enterprise-records-via-compromised-credentials-mcdonalds-vodafone-kyndryl-others

One of the hardest parts of building a cybersecurity lab is getting realistic data and incidents to practice with. So I expanded my PowerShell script. Now you can: - ⚡ Generate logs in Microsoft Sentinel - 🛡️ Automatically create a detection rule - 🚨 Simulate cybersecurity incidents - 🧪 Start practicing investigations within minutes Just download the script from GitHub, run it and you have data ready for your lab. https://github.com/Pavel-Hrabec/sentinel-logseeder-openai At the moment, I'm working on video, where I will show you step by step, how you can use it. 🙂

Today we will have 2nd community call. Everyone is welcomed. 💥 For now it doesn't have a structure, if you have any questions. Want to share something. Or just want to spend some time talking with other like minded people, make sure to join. 🔥 Last time I was working on generating logs with AI, I think today I will extend on it and produce some detection rules and incidents, which can be later used for investigation with AI. 🙂 The same script, just additional functionalities.

One thing I found interesting in OpenAI's Daybreak announcement is not just the technology itself. It is the message behind it. Attackers do not need much time anymore. If credentials, browser sessions or local secrets are exposed, the gap between compromise and impact can get very small. That changes how we should think about defense. 𝐌𝐲 𝐭𝐚𝐤𝐞𝐚𝐰𝐚𝐲𝐬: ✅ Protect identities and sessions, not only passwords ✅ Assume local devices can become the entry point to bigger incidents ✅ Reduce long-lived sessions where possible ✅ Prioritize phishing-resistant MFA for sensitive accounts ✅ Shorten detection and response time around account abuse A lot of security teams still focus heavily on perimeter alerts while identity and session theft can move much faster. I will be requesting to access Daybreak. 🙂 Can't wait to try it out. https://openai.com/index/expanding-daybreak-as-the-cyber-defense-window-narrows/

You can use these models for free now to learn cybersecurity. 💰 - gpt-4-1-mini - gpt-5-4-mini - gpt-5-6-sol - gpt-5-3-codex - Kimi-K2-7-Code - DeepSeek-V4-Pro - grok-4-3 - claude-haiku-4-5 I will add more AI models later. For now, you can use those. HTTP - https://apim-ai-lab-swc-ph-20260807.azure-api.net Content-Type: application/json Key - ffdf9b7d05054f288609e962b54cad32 I will also cover this in video and explain you, how you can use them. You shouldn't spend more money once you join community. 👊 🚨 Please use the key only for test environments. There are limits for millions of tokens each month so everyone can use it without problems, if we are all responsible.😊