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:
- Kremlin spokesman lists top countries where cyber attacks originate
June 17, 2021
Kremlin Spokesman Dmitry Peskov has prepared a list of the top countries, where cyber attacks originate, at the request of Russian President Vladimir Putin, handing over this list to reporters. “In the first half of 2020, the leaders among all countries where all types of cyber attacks originated are: the US, Canada, Brazil, Mexico, the UK,” ...
- Bash Ransomware DarkRadiation Targets Red Hat- and Debian-based Linux Distributions
June 17, 2021
A recently discovered Bash ransomware piqued our interest in multiple ways. Upon investigating, we found that the attack chain is fully implemented as a bash script, but it also seems that the scripts are still under development. Most components of this attack mainly target Red Hat and CentOS Linux distributions; however, in some scripts Debian-based Linux ...
- Matanbuchus: Malware-as-a-Service with Demonic Intentions
June 16, 2021
Unit 42 researchers often spend time investigating what we call non-traditional sources. Non-traditional sources often include underground marketplaces and sites, spanning from forums on the Tor network to Telegram channels and other marketplaces. One such case that we investigated involves a threat actor called BelialDemon, who is a member of several underground forums and marketplaces. In ...
- Ferocious Kitten: 6 years of covert surveillance in Iran
June 16, 2021
Ferocious Kitten is an APT group that since at least 2015 has been targeting Persian-speaking individuals who appear to be based in Iran. Although it has been active for a long time, the group has mostly operated under the radar and has not been covered by security researchers to the best of our knowledge. It ...
- Ransomware Poll: 80% of Victims Don’t Pay Up
June 16, 2021
Ransomware is on the rise, but what toll does it take on the real world? Threatpost set out to answer that question in an exclusive poll aimed at taking the pulse of organizations wrestling with attacks, including looking at mitigations and the defenses organizations have in place. When viewed against the backdrop of complementary reports from ...
- Smoking Out a DARKSIDE Affiliate’s Supply Chain Software Compromise
June 16, 2021
Mandiant observed DARKSIDE affiliate UNC2465 accessing at least one victim through a Trojanized software installer downloaded from a legitimate website. While this victim organization detected the intrusion, engaged Mandiant for incident response, and avoided ransomware, others may be at risk. As reported in the Mandiant post, “Shining a Light on DARKSIDE Ransomware Operations,” Mandiant Consulting has ...