Skip to main content
Mallory
← Back to News
ai-enabled-threat-activitycybercrime-service-ecosystemindustrials-sector-threatfinancial-sector-threat

Google Reports AI-Assisted Aerospace Intrusion and Stolen Enterprise AI Access Market

Emerging Technology Risk · 13 sources · Updated 1h ago · First seen Nov 16, 2026

Google Reports AI-Assisted Aerospace Intrusion and Stolen Enterprise AI Access Market

Google Threat Intelligence Group reported that attackers used Cairn, a publicly available AI-enabled penetration-testing framework associated with the Chinese open-source security community, in an intrusion targeting an unnamed U.S. aerospace organization. Google did not attribute the intrusion to a government, establish its precise timing, or disclose whether information was stolen. Its broader findings describe attackers adopting agentic frameworks to accelerate multistage intrusions, including an automated workflow spanning software-patch analysis, exploit development, and execution. Chinese actors also used AI to analyze stolen information and prepare intelligence reports. North Korean attackers targeted cryptocurrency organizations through fake recruitment approaches and meeting services, then modified Claude Code settings in company repositories to trigger malware downloads. Google also reported routine AI use by Russia-, China-, and Iran-linked influence operations to generate and distribute content.

The report also described an illicit market for “LLM Mixers,” which resell enterprise AI access obtained through stolen credentials and distribute customer queries across compromised environments to sustain access and reduce detection. Advertised token prices ranged from $0.10 to $25, while Mandiant observed incidents generating more than $100,000 in unauthorized AI usage within 24 hours and cumulative charges reaching $1 million at some organizations. Beyond financial losses, compromised access can support malicious AI deployments and potentially expose corporate data. Enterprises should enforce token and spending limits, alert on anomalous usage, secure and rotate API keys, audit accounts, strengthen phishing protections, and require multifactor authentication.

New surface, new adversariesTrack how attackers are adapting to this technology.
Track this risk
Event timeline

How this story unfolded

Events from the most recent confirmed update back to the earliest known activity.

15 events · Jun 2026 – Nov 13, 2026
  1. Reported Nov 17, 20263h ago

    Google identifies APT42 using LLMs for reconnaissance and social engineering

    Google reported that Iranian government-backed actor APT42 used several large language models, including Gemini, for targeted reconnaissance and social engineering.

    Chinese state group developing automated attack tool for cyber espionage, Google says | The Record from Recorded Future News
  2. Reported Nov 16, 20261d ago

    Google SecOps releases ARTEX detection rules

    Google SecOps released detection rules for the ARTEX automated hacking framework. The report emphasizes behavioral detection because many observed indicators originate from LLM output or invoked tools rather than fixed framework code.

    Google GTIG Warns AI Agents Automate Exploits, Steal Credentials and Scale Cybercrime
  3. Reported Nov 16, 20261d ago

    Google identifies AI-assisted influence platform built to manage over 100,000 accounts

    Google reported that an influence-operation contractor used the Gemini API while developing a platform designed to operate more than 100,000 social media accounts through cloud phone farms. The platform used AI to simulate ordinary scrolling and viewing to age fake accounts before distributing political messages.

    Google Reveals How Hackers Are Using AI Agents to Automate Real-World Cyberattacks | Cryptika Cybersecurity
  4. Reported Nov 16, 20261d ago

    Mandiant documents unauthorized AI charges reaching $1 million

    Mandiant documented incidents generating more than $100,000 in unauthorized AI usage within 24 hours. At some organizations, cumulative charges reached $1 million before access was revoked.

    Your enterprise AI access can be sold for pennies - beware 'LLM Mixers' - ZDNET
  5. Reported Nov 16, 20261d ago

    GTIG investigates illicit AI-access storefronts known as LLM Mixers

    GTIG analyzed Russian- and Chinese-language marketplace advertisements offering unauthorized access to AI services through stolen credentials and proxy infrastructure. The services distributed customer queries across compromised environments to maintain access and reduce detection.

    Your enterprise AI access can be sold for pennies - beware 'LLM Mixers' - ZDNET
  6. Reported Nov 16, 20261d ago

    North Korean attackers alter Claude Code settings to download malware

    After gaining access, the North Korean attackers modified Claude Code settings in company code repositories. Starting the assistant with those settings triggered a malware download.

    AI hacking tool used in US aerospace sector intrusion, Google says - Nextgov/FCW
  7. Reported Nov 16, 20261d ago

    North Korean fake recruiters target cryptocurrency organizations

    A North Korean group targeted cryptocurrency organizations in North America and Europe with fake recruiter profiles and fraudulent meeting services. Instructions to fix fabricated technical problems induced targets to execute malicious code.

    AI hacking tool used in US aerospace sector intrusion, Google says - Nextgov/FCW
  8. Reported Nov 16, 20261d ago

    Google identifies fully automated patch-to-exploitation workflow

    Google identified an attacker-controlled system configured to analyze software fixes, develop exploits, and execute them without human involvement. Its researchers described this as their first observation of attackers fully automating that process.

    AI hacking tool used in US aerospace sector intrusion, Google says - Nextgov/FCW
  9. Reported Nov 16, 20261d ago

    China-linked group develops an automated hacking system

    Google observed a China-linked group developing its own automated hacking system, assessed as likely intended for espionage operations.

    AI hacking tool used in US aerospace sector intrusion, Google says - Nextgov/FCW
  10. Reported Nov 16, 20261d ago

    Chinese contractor uses AI to analyze stolen information

    Google identified a Chinese contractor using AI to translate and analyze stolen information, infer targeted organizations' structures, and compile intelligence reports for customers.

    AI hacking tool used in US aerospace sector intrusion, Google says - Nextgov/FCW
  11. Reported Nov 16, 20261d ago

    Attackers deploy Cairn against a U.S. aerospace organization

    Google reported that attackers used Cairn, a publicly available AI-backed penetration-testing framework, in an intrusion targeting an unnamed U.S. aerospace organization. Google did not establish the intrusion's exact timing, attribute it to a government, or disclose whether information was stolen.

    AI hacking tool used in US aerospace sector intrusion, Google says - Nextgov/FCW
  12. Reported Nov 16, 20262d ago

    Google reports wave of nonfunctional AI-generated exploits

    Google reported a wave of LLM-generated exploits claimed to work but found to be nonfunctional. These unreliable exploits create additional verification work for defenders.

    Google Warns AI Agents Are Accelerating Cyberattacks
  13. Nov 13, 20265d ago

    FBI flags ARTEX in cybersecurity advisory

    The FBI published a cybersecurity advisory flagging ARTEX, an automated hacking framework also discussed in Google's subsequent threat intelligence findings.

    AI hacking tool used in US aerospace sector intrusion, Google says - Nextgov/FCW
  14. Aug 2026

    Attackers use customized PentAGI to scan organizations across Asia

    Attackers used a customized PentAGI platform to conduct reconnaissance and vulnerability scanning against organizations across Asia between late August and early October 2026.

    Google Reveals How Hackers Are Using AI Agents to Automate Real-World Cyberattacks
  15. Jun 2026

    Jieyou phishing-as-a-service platform integrates an LLM chatbot

    The Chinese phishing-as-a-service platform Jieyou reportedly integrated an LLM chatbot in June 2026.

    GTIG : Les frameworks multi-agents IA franchissent un seuil critique dans les cyberattaques | CyberVeille
Linked entities

Related entities

Vulnerabilities, threat actors, malware, products, organizations, and breaches Mallory has linked to this story.

Open in app →

Malware

2 linked

Affected products

1 linked
Claude Code

Organizations

2 linked
APIMartJetNight
Do you run this?Follow how adversaries are adapting to this technology, and where it touches your stack today.
Track This Risk
Mallory platform

See the full picture, correlated to your attack surface.

This page covers what’s public. Mallory adds the parts that aren’t: which of your assets are affected, which threat actors are using it right now, and what to deploy next.

Exposure mappingMap indicators from this story to your assets and identify affected systems in minutes.
Threat actor evidenceEvery observed campaign, victim, and pivot linked to actors named in this story.
Associated malwareMalware, exploits, and IOCs connected to the activity described here.
Detection signaturesYARA, Sigma, and Snort rules deployed to your SIEM as soon as they’re published.
Scheduled alertsGet matching new stories delivered to your team as they break, not the next morning.
AI threadsAsk questions about this story and take action on the answers.