BT
Privacy ToolboxJournalProjectsResumeBookmarks
Feed
Privacy Toolbox
Journal
Projects
Resume
Bookmarks
Intel
Threat Actors
Privacy Threats
Dashboard
CVEs
Tags
Intel
Threat ActorsPrivacy ThreatsDashboardCVEsTags

Intel

  • Feed
  • Threat Actors
  • Privacy Threats
  • Dashboard
  • Privacy Toolbox
  • CVEs

Personal

  • Journal
  • Projects

Resources

  • Subscribe
  • Bookmarks
  • Developers
  • Tags
Cybersecurity News & Analysis
github
defconxt
•
© 2026
•
blacktemple.net
  1. Feed
  2. /APT Roundup: Zero-Day Exploitation, Iranian Cyber Escalation, and AI-Targeted Attacks

APT Roundup: Zero-Day Exploitation, Iranian Cyber Escalation, and AI-Targeted Attacks

March 3, 2026Nation-State & APT4 min readhigh

Originally reported by Security Affairs, The Record, Palo Alto Unit 42

#apt28#apt37#iran-cyber-threats#zero-day-exploitation#ai-prompt-injection#chrome-vulnerability#air-gapped-networks#cyber-command
Share

TL;DR

APT28 exploited an MSHTML zero-day while APT37 developed new air-gapped infiltration techniques using cloud storage and USB implants. Meanwhile, Iranian cyber threats are escalating amid Middle East tensions, prompting warnings from UK and US authorities.

Why high?

Multiple actively exploited zero-day vulnerabilities by nation-state actors (APT28's MSHTML exploit, Chrome Gemini vulnerability) combined with escalating Iranian cyber threats warrant high severity classification.

APT Roundup: Zero-Day Exploitation, Iranian Cyber Escalation, and AI-Targeted Attacks

Nation-state threat actors are adapting their tactics with sophisticated new approaches targeting AI systems, air-gapped networks, and browser vulnerabilities. Recent developments span from zero-day exploitation campaigns to emerging AI prompt injection attacks, while geopolitical tensions drive increased Iranian cyber activity.

Chrome Vulnerability Enables AI Assistant Hijacking

Palo Alto Networks researchers discovered a critical Chrome vulnerability (CVE-2026-0628) that allows malicious extensions to hijack Google's Gemini Live AI assistant. The flaw enables attackers to spy on users and exfiltrate sensitive files by taking control of the AI interface. This represents a new attack vector targeting AI-powered browser features, demonstrating how threat actors are adapting to exploit emerging technologies integrated into everyday tools.

APT28 Exploits MSHTML Zero-Day Before Patch

Akamai researchers report that Russia-linked APT28 (Fancy Bear) exploited CVE-2026-21513, a high-severity MSHTML vulnerability with a CVSS score of 8.8, before Microsoft patched it in February 2026. The Internet Explorer security control bypass vulnerability can lead to code execution, highlighting APT28's continued capability to weaponize zero-day vulnerabilities for espionage operations. The pre-patch exploitation underscores the group's access to previously unknown vulnerabilities.

APT37 Develops Air-Gapped Network Infiltration Technique

North Korean group APT37 (ScarCruft) deployed new tools in the "Ruby Jumper" campaign, combining cloud storage and USB-based malware to breach air-gapped systems. Zscaler ThreatLabz discovered the group using a backdoor that leverages Zoho WorkDrive for command and control communications alongside USB-based implants designed to traverse air-gapped network boundaries. This dual-vector approach represents a significant evolution in techniques for infiltrating highly secured environments.

Iranian Cyber Threat Escalation Across Multiple Fronts

Geopolitical tensions in the Middle East have triggered widespread concern about Iranian cyber capabilities. The UK's National Cyber Security Centre (NCSC) issued warnings about potential Iranian cyberattacks, urging British organizations to strengthen defensive measures amid escalating regional conflict. Unit 42 researchers documented recent Iranian cyberattack activity including phishing campaigns, hacktivist operations, and cybercrime activities, noting a marked increase in aggressive posturing.

Simultaneously, U.S. Cyber Command confirmed conducting offensive cyber operations against Iranian communications systems and sensors, with the country's top general stating these attacks helped set conditions for joint bombing campaigns with Israel. This represents a significant escalation in cyber warfare operations targeting Iranian infrastructure.

AI Prompt Injection Attacks Observed in Wild

Unit 42 researchers documented real-world indirect prompt injection attacks where adversaries weaponize hidden web content to exploit large language models for high-impact fraud. These attacks represent a new category of threats targeting AI agents, demonstrating how malicious actors can manipulate AI systems through carefully crafted web content that appears benign to human users but triggers malicious behavior in AI processing systems.

Additional Notable Incidents

The University of Hawaii Cancer Center confirmed a data breach following a ransomware attack, with compromised information traced back to a 1993 Multiethnic Cohort Study that used driver's license numbers and voter registration records. German courts convicted an alleged mastermind behind a global investment scam network, with the Bamberg Regional Court conducting an 11-day trial under heightened security measures.

Sources

  • Chrome security flaw enabled spying via Gemini Live assistant
  • Middle east crisis prompts UK NCSC warning on potential Iranian cyber activity
  • Russia-linked APT28 exploited MSHTML zero-day CVE-2026-21513 before patch
  • APT37 combines cloud storage and USB implants to infiltrate air-gapped systems
  • Cyber Command disrupted Iranian comms, sensors, top general says
  • University of Hawaiʻi Cancer Center confirms data leak following ransomware attack
  • German court convicts alleged mastermind behind global investment scam network
  • British organizations urged to be alert to threat of Iranian cyberattacks
  • Fooling AI Agents: Web-Based Indirect Prompt Injection Observed in the Wild
  • Threat Brief: March 2026 Escalation of Cyber Risk Related to Iran

Originally reported by Security Affairs, The Record, Palo Alto Unit 42

Tags

#apt28#apt37#iran-cyber-threats#zero-day-exploitation#ai-prompt-injection#chrome-vulnerability#air-gapped-networks#cyber-command

Threat Actors

🇷🇺APT28
🇰🇵APT37

Tracked Companies

🇺🇸Google

Related Intelligence

  • Nation-State Activity Roundup: APT28 MacroMaze Campaign, MuddyWater Operations, and Mass Infrastructure Compromises

    highMar 1, 2026
  • APT28 Targets Ukrainian Forces While Nation-State Threats Persist Globally

    highMar 11, 2026
  • Nation-State Roundup: Iran-Nexus APT Targets Iraq Officials, Phobos Admin Pleads Guilty, Multi-Year Campaign Exposed

    highMar 6, 2026

Explore

  • Dashboard
  • Privacy Threats
  • Threat Actors
← Back to the feed

Previous Article

← Critical AI Tool Flaws and Supply Chain Exposure Highlight Security Challenges

Next Article

Active Android Exploit, Government-Targeted Campaigns, and OAuth Abuse Highlight March Threat Landscape →