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:
- CISA Adds Two Known Exploited Vulnerabilities to Catalog
October 20, 2022
CISA has added two vulnerabilities to its Known Exploited Vulnerabilities Catalog, based on evidence of active exploitation. These types of vulnerabilities are a frequent attack vector for malicious cyber actors and pose significant risk to the federal enterprise. Note: to view the newly added vulnerabilities in the catalog, click on the arrow in the “Date Added ...
- Attack Surface Management 2022 Midyear Review – Part 1
October 20, 2022
The digital age is an exciting time for businesses as it offers the opportunity to be more efficient and effective with how things are done. Many companies have taken this opportunity by adopting new technologies that allow them not only to improve their operations but also to create a better culture within those organizations. However, the ...
- BlueBleed: Microsoft customer data leak claimed to be ‘one of the largest’ in years
October 20, 2022
Microsoft has confirmed one of its own misconfigured cloud systems led to customer information being exposed to the internet, though it disputes the extent of the leak. In a revelation this week, Microsoft’s Security Response Center (MSRC) said it was notified by threat intelligence firm SOCRadar on September 24 about a misconfigured endpoint that exposed business ...
- CISA Releases Three Industrial Control Systems Advisories
October 20, 2022
CISA has released three (3) Industrial Control Systems (ICS) advisories on October 20, 2022. These advisories provide timely information about current security issues, vulnerabilities, and exploits surrounding ICS. CISA encourages users and administrators to review the newly released ICS advisories for technical details and mitigations: ICSA-22-293-01 Bentley Systems MicroStation Connect ICSMA-21-294-01 B Braun Infusomat Space Large Volume Pump ...
- From RM3 to LDR4: URSNIF Leaves Banking Fraud Behind
October 20, 2022
A new variant of the URSNIF malware, first observed in June 2022, marks an important milestone for the tool. Unlike previous iterations of URSNIF, this new variant, dubbed LDR4, is not a banker, but a generic backdoor (similar to the short-lived SAIGON variant), which may have been purposely built to enable operations like ransomware and ...
- CISA Updates Advisory on Threat Actors Exploiting Multiple CVEs Against Zimbra Collaboration Suite
October 19, 2022
CISA and the Multi-State Information Sharing & Analysis Center (MS-ISAC) have updated joint Cybersecurity Advisory AA22-228A: Threat Actors Exploiting Multiple CVEs Against Zimbra Collaboration Suite, originally released August 16, 2022. The advisory has been updated to reference the addition of a new Malware Analysis Report, MAR-10398871.r1.v2. CISA encourages organizations to review the latest update to AA22-228A ...