← Back to news

New Gaslight macOS Malware Uses Prompt Injection to Disrupt AI-Assisted Analysis

The Hacker News25/06/2026, 09:23
Read full article →

Summary

AI-Generated

Key Points:

  • A new Rust-based macOS malware, codenamed Gaslight, has been identified, featuring a prompt injection payload designed to disrupt AI-assisted malware analysis.
  • The malware, attributed to North Korean threat actors, uses a Telegram bot for command-and-control and can gather extensive system information while employing deceptive messages to confuse detection tools.
  • Recommended actions include monitoring for unusual LaunchAgent entries, implementing strict network controls to block unauthorized C2 communications, and enhancing AI detection capabilities to recognize prompt injection tactics.

Technical Details: Gaslight utilizes a Telegram bot API for C2 communication and embeds a Base64-encoded Python script for data exfiltration. It achieves persistence through a LaunchAgent labeled "com.apple.system.services.activity."

MITRE ATT&CK Techniques:

  • T1071.001 - Application Layer Protocol: Web Protocols (Command and Control)
  • T1053.001 - Scheduled Task/Job: LaunchAgent (Persistence)
  • T1041 - Exfiltration Over Command and Control Channel (Exfiltration)

IOCs Mentioned: None mentioned

Join the discussion — sign up to comment, upvote, and save articles.

Discussion

or to comment
Loading...

Loading comments...

Join 5,000+ security professionals

Get access to curated threat intel, upvote articles, join discussions, and build your karma in the SOC community.