42.6% of MCP server repos with confirmed exploitable vulnerabilities show signs of AI-generated code.
With Alfredo Oliveira at TrendAI Research, we swept 19,000+ MCP servers; analysis shows 600–1,650 repos contain exploitable vulnerabilities.
https://t.co/bk5DrPKgcZ
MCP adoption is exploding as AI agents plug into everything, but security is lagging badly. Exposed MCP servers nearly tripled to 1,467 in just months, many running with zero authentication and leaking direct access to cloud services. Read more: https://t.co/HFiWj7Uxki
Heads up for #RSAC2026 attendees — at #CloudVillage, Alfredo Oliveira (Sr. Security Researcher, TrendAI) is presenting: "Mitigating MCP Server Threats in Cloud Environments" Worth the stop by 1:50PM Schedule: https://t.co/J6dvNLNi5c #MCPSecurity#CloudSecurity#TrendAI
@SalamonTomas Význam to nemění ;) AI rozhodně mění pravidla hry.
A používání "out of the box" řešení je risk a je jedno, jestli je to docker image nebo python package.
AI jde použít i na obranu, ale pro použitelnost je to komplexnější problém než 1 prompt. např. zveličování problémů,etc.
The TrendAI™ State of AI Security Report is out.
As the AI threat landscape evolves, organizations must adopt proactive strategies to address unique risks from supply chain vulnerabilities and exposed infrastructure.
Strengthen your AI defenses: https://t.co/pc3rr92q05
Beware of AI applications and their privacy policy claims. In new research, we reveal when security measures fail miserably, resulting in the exposure of sensitive user photos and proprietary AI models due to unsecured cloud storage.
https://t.co/wLRh7HYx2Q
As part of our continuous research on how to strengthen Model Context Protocol (MCP) security we explore how proper MCP server hardening within containers can mitigate security risks and offer security benefits essential for protecting cloud workloads. https://t.co/BicSZU8sQQ
Don't be a victim of fast and unsafe MCP server deployments. MCP Security: Network-Exposed Servers Are Backdoors to Your Private Data https://t.co/zi6c7raEnd
Organizations running AI-powered services with proprietary models could be at risk of intellectual property theft, particularly if their models or inference services are exposed through misconfigured APIs.
Here’s what you need to know: https://t.co/96wRpcUM2Y
The exposure of OpenVPN and SSH keys within a container image can lead to full network compromise. Learn how to protect your keys and certificates today.
Read more here:⬇️ https://t.co/LNaSqOzLbe
Cybercriminals can weaponize stolen certificates and private keys to infiltrate systems undetected. Learn how to safeguard your organization from these risks.
Here’s what you need to know: ⬇️ https://t.co/LNaSqOzLbe