New Prompt Injection Attack Vectors Through MCP Sampling


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:

  • Adversaries Leverage AI for Vulnerability Exploitation, Augmented Operations, and Initial Access

    May 11, 2026

    Since our February 2026 report on AI-related threat activity, Google Threat Intelligence Group (GTIG) has continued to track a maturing transition from nascent AI-enabled operations to the industrial-scale application of generative models within adversarial workflows. This report, based on insights derived from Mandiant incident response engagements, Gemini, and GTIG’s proactive research, highlights the dual nature ...

  • Experts warn nearly half of the world’s passwords can easily be cracked in just a few minutes

    May 8, 2026

    Using real-world samples recovered from the dark web, Kaspersky researchers have tested how long it would take to crack most passwords, and found that almost half of the world’s passwords can be cracked in less than a minute. Additionally, the research shows that within an hour, that number rises to three out of five passwords. Armed with this knowledge, ...

  • Disgraced US gov software contractor found guilty of database destruction

    May 8, 2026

    A Virginia man, Sohaib Akhter, faces decades in prison after a jury convicted him of being involved in a scheme to delete approximately 96 databases containing US government data. The events of the case transpired around two weeks before the twin brothers allegedly involved were fired from their jobs at a software supplier to the US ...

  • Poland says hackers breached water treatment plants, and the US is facing the same threat

    May 8, 2026

    Poland’s intelligence service said it detected attacks on five water treatment plants where hackers could have taken control of the industrial equipment inside, including, in the worst case, tampering with the safety of the water supply. The story is relevant beyond Poland’s borders: U.S. water infrastructure has faced similar threats in recent years. In 2021, a ...

  • Worm rubs out competitor’s malware, then takes control

    May 8, 2026

    There’s a mysterious framework worming its way through exposed cloud instances removing all traces of TeamPCP infections, but it’s not benevolent by a long shot: Whoever is behind this bit of malware may be cleaning up who came before, but only so they can take their place. Discovered by security outfit SentinelOne’s SentinelLabs researchers and dubbed PCPJack ...

  • ‘Dirty Frag’ Linux flaw one-ups CopyFail with no patches and public root exploit

    May 8, 2026

    Broken disclosure embargo left admins facing a fresh root-level flaw with no CVE A fresh Linux privilege escalation bug dubbed “Dirty Frag” has dropped into the wild with no patches, no CVE, and a public exploit that hands attackers root access across major distributions.Security researcher Hyunwoo Kim disclosed the local privilege escalation flaw on Friday after what he ...