HAIJA INTEL REPORT

I wanted to address the speculation about the recently introduced Device Bound Session Credentials (DBSC) security feature in Google Chrome. Does it help increase the security of session cookies against infostealer malware and MFA phishing? The feature has been available and Goog

I've published an article on how an attacker could break into the Claude Code supply chain! We've published a new blog post by RyotaK @ryotkak. He discovered a vulnerability that allows attackers to bypass the permission controls of Claude Code GitHub Actions via a GitHub issue,

During an assessment, our team discovered that StrongDM auth state files containing JWTs & key material could be reused across hosts to obtain authenticated sessions & access infrastructure resources (CVE-2026-4387). Re…

I'm excited to be able to finally publish the public disclosure for CVE-2026-4387. Check out my blog on discovering the reuse of the state.kv file to get authenticated sessions with StrongDM (now fixed).

Attending #CiscoLive in Las Vegas? Stop by booth 2206 and connect with our team. We'd love to discuss your identity attack path management questions and show you how BloodHound Enterprise helps organizations uncover and reduce hidden risk.

Tomorrow we're live at 1 p.m. ET / 10 a.m. PT for our latest episode of SecOps Weekly! Phil Hagen and Chris Brook are hopping on to chat about the latest security trends and answer audience questions from our mailbag. Join us live to hear their take and learn what you and your

Don't miss "Signal to Enforcement: A Coordinated Security Walkthrough" featuring @jaredcatkinson at 4:30 PM. Follow a real-world attacker path across identity, SaaS, devices, and networks, and see how integrated security architectures enable coordinated response and enforcement.

I found a $3,000 bug in an AI chatbot using prompt injection. Video https:// youtu.be/Q6hQlM6f6Cs Lab https:// app.hackinghub.io/hubs/shopmate

Your “agentic coworker” is sending suspicious messages via Microsoft Teams. It’s going to need to have a chat with the agentic HR department. Read Part 2 of our series on investigating suspicious AI workflows in Entra ID: https:// redcanary.com/blog/threat-de tection/entra-id-ai-

The problem with agentic security workflows isn't access to data. It's making that data callable from the agent. The Sysdig MCP server on AWS Marketplace closes that gap, inside Amazon Bedrock AgentCore. Read the blog: https:// okt.to/uVP2Fe #CloudSecurity

As I'm sure you've all seen by now, nerds have been exploiting Meta's AI agent goop to steal Instagram accounts. The Instagram AI agent for support could be convinced to reset the credentials to other users accounts by asking nicely and do a super gnarly kickflip on a