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
β€’
Β© 2026
β€’
blacktemple.net
  1. Feed
  2. /AI Weaponization and Major Data Breaches Dominate Cybercrime Landscape

AI Weaponization and Major Data Breaches Dominate Cybercrime Landscape

March 2, 2026Nation-State & APT3 min readhigh

Originally reported by Security Affairs

#data-breach#ai-abuse#cybercrime#telecom#government#fake-ids#vulnerability#law-enforcement
Share

TL;DR

Major cybercrime developments include ShinyHunters releasing the full Odido dataset in what's being called the Netherlands' biggest data leak in history, and attackers using Anthropic's Claude Code AI to steal 150GB from Mexican government systems. Meanwhile, law enforcement operations resulted in 30 arrests targeting child exploitation networks and the takedown of an AI-powered fake ID operation.

Why high?

The Netherlands faces its biggest data leak in history with the Odido breach, while AI tools are being weaponized for sophisticated attacks against government systems.

The cybercrime landscape has seen significant developments this week, spanning from AI weaponization in government attacks to massive telecommunications breaches and coordinated law enforcement operations.

ShinyHunters Releases Complete Odido Dataset in Historic Dutch Breach

Cybercrime group ShinyHunters has leaked the complete dataset from Odido, the Dutch telecommunications company formed from the 2023 rebrand of T-Mobile Netherlands and Tele2. Security researchers are calling this the biggest data leak in Netherlands history.

Odido operates as one of the largest mobile network operators in the Netherlands, serving millions of customers. The breach potentially exposes sensitive customer data including personal information, account details, and telecommunications metadata. The full scope of the compromised data remains under investigation by Dutch authorities.

Claude AI Weaponized in 150GB Mexican Government Data Theft

Attackers leveraged Anthropic's Claude Code AI assistant to orchestrate a sophisticated cyberattack against Mexican government systems, according to Israeli cybersecurity firm Gambit Security. The threat actors used the AI tool to develop custom exploits, create specialized attack tools, and automate the exfiltration of over 150GB of sensitive government data.

This incident represents a concerning evolution in threat actor capabilities, demonstrating how publicly available AI coding assistants can be repurposed for large-scale cyber espionage operations. The automated nature of the attack suggests a level of sophistication that could lower barriers for less technically skilled attackers.

Europol's Project Compass Dismantles Child Exploitation Network

Europol's year-long Project Compass operation resulted in 30 arrests targeting "The Com," a cybercrime network specializing in crimes against children and teenagers. The coordinated international effort identified 62 victims and successfully protected four children from immediate harm.

The operation demonstrates continued law enforcement focus on networks exploiting minors online. Project Compass involved multiple European agencies working in coordination to map the network's infrastructure and identify both operators and victims across jurisdictions.

Ukrainian Pleads Guilty in OnlyFake AI-Generated ID Operation

Ukrainian citizen Yurii Nazarenko pleaded guilty to operating OnlyFake, an AI-powered platform that generated and sold over 10,000 counterfeit identity documents globally. The operation, prosecuted by the Southern District of New York, represents one of the largest AI-driven identity fraud schemes uncovered to date.

OnlyFake utilized artificial intelligence to create realistic-looking government identification documents, bypassing traditional detection methods used by financial institutions and other verification systems. The case highlights the dual-use nature of AI technology in criminal enterprises.

ClawJacked Vulnerability Patched in OpenClaw AI Agent Platform

Oasis Security researchers disclosed "ClawJacked," a high-severity vulnerability in the OpenClaw AI agent platform that allowed malicious websites to hijack local AI agent instances through brute-force attacks. The flaw enabled silent data theft from compromised systems.

OpenClaw addressed the vulnerability in version 2026.2.26, released following coordinated disclosure. The issue underscores emerging security challenges in AI agent architectures, particularly around authentication and access control for locally-running AI services.

Sources

  • Europol's Project Compass nets 30 arrests in crackdown on "The Com"
  • ClawJacked flaw exposed OpenClaw users to data theft
  • Ukrainian hacker pleads guilty to running OnlyFake AI ID scam site
  • ShinyHunters leaked the full Odido dataset
  • Claude code abused to steal 150GB in cyberattack on Mexican agencies

Originally reported by Security Affairs

Tags

#data-breach#ai-abuse#cybercrime#telecom#government#fake-ids#vulnerability#law-enforcement

Threat Actors

🏴ShinyHunters

Tracked Companies

πŸ‡ΊπŸ‡ΈT-Mobile

Related Intelligence

  • Russian APT Deploys New Ukraine-Targeting Malware as Law Enforcement Disrupts Global Cybercrime Operations

    mediumMar 5, 2026
  • Nation-State Activity Roundup: APT28 MacroMaze Campaign, MuddyWater Operations, and Mass Infrastructure Compromises

    highMar 1, 2026
  • Iranian Actors Hit Medical Infrastructure While Meta Disrupts Influence Operations

    highMar 12, 2026

Explore

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

Previous Article

← APT28 Exploits MSHTML Zero-Day; North Korean npm Campaign Evolves

Next Article

Weekly Security Roundup: Teramind Phishing Campaign Targets Remote Workers β†’