The jailbreak technique “Bad Likert Judge” manipulates LLMs to generate harmful content using Likert scales, exposing safety gaps in LLM guardrails. The post Bad Likert Judge: A Novel Multi-Turn Technique to Jailbreak LLMs by Misusing Their Evaluation Capability appeared first…
Tag: Unit 42
Now You See Me, Now You Don’t: Using LLMs to Obfuscate Malicious JavaScript
This article demonstrates how AI can be used to modify and help detect JavaScript malware. We boosted our detection rates 10% with retraining. The post Now You See Me, Now You Don’t: Using LLMs to Obfuscate Malicious JavaScript appeared first…
Effective Phishing Campaign Targeting European Companies and Organizations
A phishing campaign targeting European companies used fake forms made with HubSpot’s Free Form Builder, leading to credential harvesting and Azure account takeover. The post Effective Phishing Campaign Targeting European Companies and Organizations appeared first on Unit 42. This article…
Effective Phishing Campaign Targeting European Companies and Institutions
A phishing campaign targeting European companies used fake forms made with HubSpot’s Free Form Builder, leading to credential harvesting and Azure account takeover. The post Effective Phishing Campaign Targeting European Companies and Institutions appeared first on Unit 42. This article…
LDAP Enumeration: Unveiling the Double-Edged Sword of Active Directory
Using real-world examples and offering plenty of pragmatic tips, learn how to protect your directory services from LDAP-based attacks. The post LDAP Enumeration: Unveiling the Double-Edged Sword of Active Directory appeared first on Unit 42. This article has been indexed…
Dirty DAG: New Vulnerabilities in Azure Data Factory’s Apache Airflow Integration
Vulnerabilities in Microsoft Azure Data Factory’s integration with Apache Airflow can lead to unauthorized access and control over cloud resources. The post Dirty DAG: New Vulnerabilities in Azure Data Factory’s Apache Airflow Integration appeared first on Unit 42. This article…
Crypted Hearts: Exposing the HeartCrypt Packer-as-a-Service Operation
Analysis of packer-as-a-service (PaaS) HeartCrypt reveals its use in over 2k malicious payloads across 45 malware families since its early 2024 appearance. The post Crypted Hearts: Exposing the HeartCrypt Packer-as-a-Service Operation appeared first on Unit 42. This article has been…
Network Abuses Leveraging High-Profile Events: Suspicious Domain Registrations and Other Scams
Unit 42 probes network abuses around events like the Olympics, featuring case studies of scams and phishing through domain registrations and more. The post Network Abuses Leveraging High-Profile Events: Suspicious Domain Registrations and Other Scams appeared first on Unit 42.…
Threat Assessment: Howling Scorpius (Akira Ransomware)
Howling Scorpius, active since 2023, uses Akira ransomware to target businesses globally, employing a double-extortion strategy and upgrading tools regularly. The post Threat Assessment: Howling Scorpius (Akira Ransomware) appeared first on Unit 42. This article has been indexed from Unit…
Fake North Korean IT Worker Linked to BeaverTail Video Conference App Phishing Attack
North Korean IT worker cluster CL-STA-0237 instigated phishing attacks via video apps in Laos, exploiting U.S. IT firms and major tech identities. The post Fake North Korean IT Worker Linked to BeaverTail Video Conference App Phishing Attack appeared first on…
Global Companies Are Unknowingly Paying North Koreans: Here’s How to Catch Them
We discuss North Korea’s use of IT workers to infiltrate companies, detailing detection strategies like IT asset management and IP analysis to counter this. The post Global Companies Are Unknowingly Paying North Koreans: Here’s How to Catch Them appeared first…
ModeLeak: Privilege Escalation to LLM Model Exfiltration in Vertex AI
New research reveals two vulnerabilities in Google’s Vertex AI that may lead to privilege escalation or data theft through custom jobs or malicious models. The post ModeLeak: Privilege Escalation to LLM Model Exfiltration in Vertex AI appeared first on Unit…
Silent Skimmer Gets Loud (Again)
We discuss a new campaign from the cybercrime group behind Silent Skimmer, showcasing the exploit of Telerik UI vulnerabilities and malware like RingQ loader. The post Silent Skimmer Gets Loud (Again) appeared first on Unit 42. This article has been…
Automatically Detecting DNS Hijacking in Passive DNS
Explore how we detect DNS hijacking by analyzing millions of DNS records daily, using machine learning to identify redirect attempts to malicious servers. The post Automatically Detecting DNS Hijacking in Passive DNS appeared first on Unit 42. This article has…
TA Phone Home: EDR Evasion Testing Reveals Extortion Actor’s Toolkit
A threat actor attempted to use an AV/EDR bypass tool in an extortion attempt. Instead, the tool provided Unit 42 insight into the threat actor. The post TA Phone Home: EDR Evasion Testing Reveals Extortion Actor's Toolkit appeared first on…
Jumpy Pisces Engages in Play Ransomware
A first-ever collaboration between DPRK-based Jumpy Pisces and Play ransomware signals a possible shift in tactics. The post Jumpy Pisces Engages in Play Ransomware appeared first on Unit 42. This article has been indexed from Unit 42 Read the original…
Deceptive Delight: Jailbreak LLMs Through Camouflage and Distraction
We examine an LLM jailbreaking technique called “Deceptive Delight,” a technique that mixes harmful topics with benign ones to trick AIs, with a high success rate. The post Deceptive Delight: Jailbreak LLMs Through Camouflage and Distraction appeared first on Unit…
Gatekeeper Bypass: Uncovering Weaknesses in a macOS Security Mechanism
Explore how macOS Gatekeeper’s security could be compromised by third-party apps not enforcing quarantine attributes effectively. The post Gatekeeper Bypass: Uncovering Weaknesses in a macOS Security Mechanism appeared first on Unit 42. This article has been indexed from Unit 42…
Unit 42 Looks Toward the Threat Frontier: Preparing for Emerging AI Risks
The Unit 42 Threat Frontier report discusses GenAI’s impact on cybersecurity, emphasizing the need for AI-specific defenses and proactive security. The post Unit 42 Looks Toward the Threat Frontier: Preparing for Emerging AI Risks appeared first on Unit 42. This…
Lynx Ransomware: A Rebranding of INC Ransomware
Discover recent attacks using Lynx ransomware, a rebrand of INC, targeting multiple crucial sectors in the U.S. and UK with prevalent double-extortion tactics. The post Lynx Ransomware: A Rebranding of INC Ransomware appeared first on Unit 42. This article has…