Claude Mythos is actually scary

https://www.youtube.com/watch?v=LZAZvm34rYs

#aisecurity #vulnerability #exploit

Cracks in the Bedrock: Escaping the AWS AgentCore Sandbox

https://unit42.paloaltonetworks.com/bypass-of-aws-sandbox-network-isolation-mode/

Read on HackerWorkspace: https://hackerworkspace.com/article/cracks-in-the-bedrock-escaping-the-aws-agentcore-sandbox

#malware #aisecurity #vulnerability

Anthropic pointed Claude Code at Linux kernel source files one at a time, framed as a security puzzle. It found a heap overflow in NFS code hiding since March 2003. Four more kernel bugs followed. 500+ validated vulnerabilities in weeks. Linux Foundation set aside $12.5M to help maintainers cope. Nobody found a volunteer to maintain a Google Drive library for 3.5 years. The bottleneck was never the bugs.

#LinuxKernel #OpenSource #SecurityResearch #AISecurity

🚨 NEWS 🚨

The Apache Software Foundation Launches $10M Responsible AI Initiative with Initial $1.75M Donation; Anthropic and Alpha-Omega seed fund the effort https://news.apache.org/foundation/entry/the-apache-software-foundation-launches-10m-responsible-ai-initiative-with-initial-1-75m-donation

#AIworkloads #artificialintelligence #opensource #AIsecurity

Hacking AI Agents

https://pwn.guide/free/web/hacking-ai

Read on HackerWorkspace: https://hackerworkspace.com/article/hacking-ai-agents

#cybersecurity #aisecurity #vulnerability

Another talk announcement for BSides Luxembourg!

🤖🔐 𝗔𝗜 𝗔𝗡𝗗 𝗖𝗥𝗬𝗣𝗧𝗢𝗚𝗥𝗔𝗣𝗛𝗬 𝗙𝗢𝗥 𝗘𝗩𝗔𝗦𝗜𝗩𝗘 𝗠𝗔𝗟𝗪𝗔𝗥𝗘 – zhassulan zhussupov aka @cocomelonckz 🧬🔥

Modern malware doesn’t just hide—it adapts.

This talk explores how AI and advanced cryptography are reshaping offensive tradecraft, enabling malware to rewrite itself, adapt to environments, and evade behavioral detection. From polymorphic code to stealthy encryption techniques, this is the next evolution of “thinking” malware.

zhassulan zhussupov aka @cocomelonckz is a cybersecurity researcher, author, and speaker known for deep expertise in malware development, reverse engineering, and offensive security, with multiple published works and global conference talks.

📅 Conference Dates: 6–8 May 2026 | 09:00–18:00
📍 14, Porte de France, Esch-sur-Alzette, Luxembourg
🎟️ Tickets: https://2026.bsides.lu/tickets/

📅 Schedule Link: https://pretalx.com/bsidesluxembourg-2026/schedule/

#BSidesLuxembourg2026 #AISecurity #Malware #Cryptography #RedTeam #CyberSecurity

cedric (@cedric_chee)

Project Glasswing이 프런티어 AI의 사이버 역량을 단순한 마케팅이 아닌 실질적 기술과 책임 있는 범위 설정이 있는 듀얼유즈 인프라로 다루는 새로운 단계의 신호라는 평가를 담고 있다. AI 사이버 방어와 관련한 중요한 기술적 흐름을 시사한다.

https://x.com/cedric_chee/status/2041746011459297624

#aisecurity #cybersecurity #dualuse #frontierai

Engadget (@engadget)

Anthropic이 AI를 활용해 AI 기반 साइबर 공격을 막기 위한 ‘Project Glasswing’을 공개했다. AI 보안 방어를 강화하려는 새로운 연구·프로젝트로, 공격 탐지와 대응 자동화 측면에서 의미 있는 발표다.

https://x.com/engadget/status/2041635087825920152

#anthropic #cybersecurity #aisecurity #projectglasswing #ai

Unit 42 Research Exposes Risks in Amazon Bedrock's Multi-Agent AI Systems

Unit 42's latest research reveals a hidden threat: multi-agent AI systems on Amazon Bedrock can be vulnerable to new and alarming risks, including prompt injection attacks that practitioners can't afford to ignore. Learn how to safeguard your AI applications from these emerging threats.

https://osintsights.com/unit-42-research-exposes-risks-in-amazon-bedrocks-multi-agent-ai-systems

#AmazonBedrock #MultiagentAiSystems #Unit42 #AiSecurity #PromptInjection