Ghostcommit is a proof of concept that shows how AI assistants used to review software code can be tricked by hidden instructions embedded in images.
The academic ASSET Research Group showed that an attacker can place instructions inside an image file, point to it in an AGENTS.md file, and get an AI coding agent to follow those instructions during a later task.
A pull request is basically a formal “please review and add my changes” request that a developer sends before changes are added to the main version of a software project. Human reviewers and, increasingly, AI coding tools may review the changes before they’re accepted.
Read more…
Source: MalwareBytes Labs
Sign up for the Cyber Security Review Newsletter
The latest cyber security news and insights delivered right to your inbox
Related:
- Garmin obtains decryption key after ransomware attack
July 27, 2020
Smartwatch maker Garmin has obtained the decryption key to recover its computer files from a ransomware attack last Thursday, Sky News has learned. Last week, Garmin’s services were taken offline after hackers infected the company’s networks with a ransomware virus known as WastedLocker. A number of the company’s services are operational again and the business has now ...
- Kubernetes Vulnerability Puts Clusters at Risk of Takeover (CVE-2020-8558)
July 27, 2020
A security issue assigned CVE-2020-8558 was recently discovered in the kube-proxy, a networking component running on Kubernetes nodes. The issue exposed internal services of Kubernetes nodes, often run without authentication. On certain Kubernetes deployments, this could have exposed the api-server, allowing an unauthenticated attacker to gain complete control over the cluster. An attacker with this ...
- Ensiko: A Webshell With Ransomware Capabilities
July 27, 2020
Ensiko is a PHP web shell with ransomware capabilities that targets various platforms such as Linux, Windows, macOS, or any other platform that has PHP installed. The malware has the capability to remotely control the system and accept commands to perform malicious activities on the infected machine. It can also execute shell commands on an infected ...
- NSA Urgently Warns on Industrial Cyberattacks, Triconex Critical Bug
July 24, 2020
The U.S. National Security Agency (NSA) and the Cybersecurity and Infrastructure Security Agency (CISA) have issued an alert warning that adversaries could be targeting critical infrastructure across the U.S. Separately, ICS-CERT issued an advisory on a critical security bug in the Schneider Electric Triconex TriStation and Tricon Communication Module. These safety instrumented system (SIS) controllers are ...
- OilRig Targets Middle Eastern Telecom Organization and Adds Novel C2 Channel with Steganography to Its Inventory
July 22, 2020
While analyzing an attack against a Middle Eastern telecommunications organization, Unit 42 has discovered a variant of an OilRig-associated tool we call RDAT using a novel email-based command and control (C2) channel that relied on a technique known as steganography to hide commands and data within bitmap images attached to emails. In May 2020, Symantec published ...
- MATA: Multi-platform targeted malware framework
July 22, 2020
As the IT and OT environment becomes more complex, adversaries are quick to adapt their attack strategy. For example, as users’ work environments diversify, adversaries are busy acquiring the TTPs to infiltrate systems. Recently, we reported to our Threat Intelligence Portal customers a similar malware framework that internally we called MATA. The MATA malware framework ...