This article examines the security implications of the Model Context Protocol (MCP) sampling feature in the context of a widely used coding copilot application.
MCP is a standard for connecting large language model (LLM) applications to external data sources and tools. We show that, without proper safeguards, malicious MCP servers can exploit the sampling feature for a range of attacks. We demonstrate these risks in practice through three proof-of-concept (PoC) examples conducted within the coding copilot, and discuss strategies for effective prevention.
Read more…
Source: Palo Alto Unit 42
Sign up for the Cyber Security Review Newsletter
The latest cyber security news and insights delivered right to your inbox
Related:
- Peruvian Peaks: The digital loan illusion
January 21, 2026
Crossing the Andes, we found ourselves in the digital valleys of Peru, where a new variation of the loan scam awaited us. Much like the schemes in Brazil, these operations played on hope and desperation, luring victims with promises of financial relief. The setup was so convincing that it seemed like help was just within ...
- From Extension to Infection: An In-Depth Analysis of the Evelyn Stealer Campaign Targeting Software Developers
January 19, 2026
On December 8, 2025, Koi.ai published their findings about a campaign specifically targeting software developers through weaponized Visual Studio Code extensions. Here, Trend Micro will provide a more in-depth analysis of the multistage delivery of the Evelyn information stealer. Evelyn implements multiple anti-analysis techniques to evade detection in research and sandbox environments. It collects system information ...
- StealC malware control panels could give experts the tools they need to spy on hackers
January 19, 2026
Cybersecurity researchers have managed to break into the web-based control panel for the StealC infostealer and gain valuable information on how the malware operates, and who both the attackers and the victims are. StealC is an immensely popular infostealer malware which first emerged a couple of years ago, and has since become one of the staples ...
- Firefox joins Chrome and Edge as sleeper extensions spy on users
January 19, 2026
A group of cybercriminals called DarkSpectre is believed to be behind three campaigns spread by malicious browser extensions: ShadyPanda, GhostPoster, and Zoom Stealer. Malwarebytes Labs wrote about the ShadyPanda campaign in December 2025, warning users that extensions which had behaved normally for years suddenly went rogue. After a malicious update, these extensions were able to track ...
- Newely discovered AMD CPU flaw highlights the risk of running multiple VMs
January 16, 2026
A newly discovered vulnerability in AMD chips allows malicious actors to perform remote code execution (RCE) and privilege escalation in virtual machines. Cybersecurity researchers from the CISPA Helmholtz Center for Information Security in Germany detailed a vulnerability they named StackWarp, a hardware vulnerability in AMD CPUs that breaks the protections of confidential virtual machines, by manipulating ...
- RondoDox botnet linked to large-scale exploit of critical HPE OneView bug
January 16, 2026
A critical HPE OneView flaw is now being exploited at scale, with Check Point tying mass, automated attacks to the RondoDox botnet. The security outfit says it has identified “large-scale exploitation” of CVE-2025-37164, a maximum-severity remote code execution bug in HPE’s data center management platform. Check Point has tied the activity to RondoDox, a Linux-based botnet ...