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:
- Medibank now says hackers accessed all its customers’ personal data
October 27, 2022
Australian insurance firm Medibank has confirmed that hackers accessed all of its customers’ personal data and a large amount of health claims data during a recent ransomware attack. In an announcement published today, the companies warned that an internal investigation into the attack has shown that the threat actors had far greater access to customer data ...
- Feds accuse Ukrainian of renting out PC-raiding Raccoon malware to fiends
October 26, 2022
Mark Sokolovsky, 26, a Ukrainian national, is being held in the Netherlands while he awaits extradition to America on cybercrime charges, the US Justice Department said on Tuesday. Sokolovsky, said to have used the online names Photix, Raccoon Stealer, and black21jack77777, was indicted on November 2, 2021 by a federal grand jury for his alleged role ...
- Insider Threat: The Dangers Within
October 25, 2022
Mandiant has adopted the Cyber and Infrastructure Security Agency (CISA) definition of insider, which states, “An insider is any person who has or had authorized access to or knowledge of an organization’s resources, including personnel, facilities, information, equipment, networks, and systems.” An insider threat is then the “potential for that insider to use their authorized access ...
- Hive claims ransomware attack on Tata Power, begins leaking data
October 25, 2022
Hive ransomware group has claimed responsibility for a cyber attack disclosed by Tata Power this month. A subsidiary of the multinational conglomerate Tata Group, Tata Power is India’s largest integrated power company based in Mumbai. In screenshots seen by BleepingComputer, Hive operators have posted data they claim to have stolen from Tata Power, indicating that the ransom ...
- CISA Adds Six Known Exploited Vulnerabilities to Catalog
October 24, 2022
CISA has added six 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 ...
- Apple fixes new zero-day used in attacks against iPhones, iPads
October 24, 2022
In security updates released on Monday, Apple has fixed the ninth zero-day vulnerability used in attacks against iPhones since the start of the year. Apple revealed in an advisory today that it’s aware of reports saying the security flaw “may have been actively exploited.” The bug (CVE-2022-42827) is an out-of-bounds write issue reported to Apple by an ...