AI just quietly took a MASSIVE step into cybersecurity. Anthropic’s new Project Glasswing uses its Claude Mythos Preview “superhacker” model to help big tech and critical infrastructure hunt bugs before attackers do – backed by $100M in AI credits and 40+ security partners.
CYBERSECURITY
-
Security maintainer breaks something in codebase again
By
–
Sounds like one of the security focussed maintainers broke sth again.
-
Malicious Intermediary Attacks on LLM Agent Supply Chain Security
By
–
"Your Agent Is Mine: Measuring Malicious Intermediary Attacks on the LLM Supply Chain" The danger of agent security can also exist in the API router, that is between the agent and the provider. As these routers can read prompts, keys, and tool calls in plaintext, even rewrite
-
DeepAgents introduces filesystem permissions for secure resource management
By
–
🔒 new in deepagents: filesystem permissions shared resources and org-wide policies are exactly the kind of files you want your agent to read but never overwrite. filesystem permissions let you enforce that with simple declarative allow/deny rules!
→ View original post on X — @langchain, 2026-04-13 19:16 UTC
-
Commvault AI Capabilities for Resilient Data Management Operations
By
–
See post below about @Commvault's new AI capabilities. Learn a lot more about their AI-powered #ResOps (Unified Resilience Operations), Cyber Resilience, and Identity Resilience products, services, and partners at Google Cloud Next in Las Vegas next week. Look here: commvault.com/events/google-… #Commvault #CyberResilience #IdentityResilience Commvault (@Commvault) Commvault is introducing new AI capabilities to help organizations adopt AI while maintaining control over data, agents, and recovery. From data activation to agent governance and recovery, these innovations support AI resilience across Commvault Cloud. bit.ly/3Oidvxg — https://nitter.net/Commvault/status/2043671872479301961#m
→ View original post on X — @kirkdborne, 2026-04-13 19:12 UTC
-
Exploit Costs More Than Human Security Researchers
By
–
-
Google disables accounts using CLI in headless mode
By
–
Careful, I seen reports of Google disabling accounts when their cli is used in headless mode.
-
Helping with DDoS Issues: Installation, Fixing and Verification
By
–
I’m being DDoSed with issues – best way to help is if you choose hackable install, fix it, verify it and then ping me.
-
Claude Mythos Hacking Capabilities and Security Implications
By
–
This is yet another example of Claude Mythos’s incredible hacking capabilities. I expect we’ll see more examples and independent evaluations in the coming weeks that make clear just how powerful (and dangerous, in the wrong hands) this model could be.
-
Institutional skepticism toward emerging cybersecurity threats
By
–
I largely agree with this assessment, but I haver serious misgivings about the following part: “Mythos seems to be taken seriously at a lot of large institutions and organizations filled with smart people who would rather not be worried about a new cybersecurity risk” As I’ve