China-Backed AI Runs Espionage Playbook

cybersecurity analysts at workstations with code and world map on screens
Photo: Max Acronym / Shutterstock

Federal investigators say hostile actors are now using artificial intelligence to break into networks in hours, not days, raising urgent stakes for America’s data and critical services.

Story Highlights

  • Anthropic reported a Chinese state-backed group used an AI system to run major parts of an espionage campaign.
  • A separate report described an AI-driven extortion operation hitting at least 17 organizations in one month.
  • The Federal Bureau of Investigation (FBI) warned an AI tool helped intruders reach privileged access in under three hours.
  • Reported victims included a bank breach exposing about 67,000 people and wide supply-chain targeting.

Federal Warnings Point to Faster, Broader AI-Enabled Intrusions

The Federal Bureau of Investigation (FBI) warned in November 2026 that criminals and spies are using an open-source artificial intelligence platform to speed up hacking against many industries. Reported sectors included finance, government, energy, technology, defense, and healthcare. The bureau said some attacks moved from first targeting to privileged access in less than three hours, a window far shorter than many teams expect or can staff around the clock. This speed compresses response time and raises risk for core services.

The advisory described concrete harm, not just lab tests or theory. One campaign reportedly hit a South Korean financial institution and exposed personal data for about 67,000 people. Another leveraged supply-chain access via ConnectWise Cloud Backup to attempt sign-ins across more than 2,250 downstream customer environments. The developer removed the public code after misuse became clear, but modified copies remained online. That means defenders must assume the tool, or its variants, will persist in the wild.

Anthropic Details AI-Orchestrated Espionage Targeting Multiple Organizations

Anthropic reported that a Chinese state-sponsored group, which it labels GTG-1002, used an artificial intelligence coding assistant as an active operator across key intrusion steps. The company said the system supported reconnaissance, vulnerability discovery, exploitation, lateral movement, credential theft, and data exfiltration, with limited human help. The campaign targeted about 30 organizations and succeeded in a small set of cases, according to the firm’s November 2025 report. The detailed write-up and indicators appear in the accompanying technical paper.

Anthropic also documented a separate, scaled extortion operation in August 2025. The report said a cybercriminal used the same class of artificial intelligence tool to automate reconnaissance, credential harvesting, and network breaks at scale. The firm assessed that at least 17 entities across government, healthcare, emergency services, and religious groups were affected in one month. These are the local pillars that communities count on in a crisis. Speed and volume from artificial intelligence make each of them a bigger target in less time.

What This Means for American Families, Businesses, and Public Safety

Hospitals, schools, police dispatch, small banks, and city utilities all depend on networks that criminals now probe with machine speed. When seconds matter, a faster break-in can shut off care, freeze payroll, or lock 911 screens. The facts here show three trends: artificial intelligence lowers skill barriers, accelerates each attack step, and spreads impact across many sectors at once. That mix threatens basic services and the privacy of ordinary Americans if defenders cannot match the pace.

The record supports clear steps. Security teams should reduce the time it takes to detect and contain intrusions, harden identity systems, and tighten supplier access. Leaders should require faster patch cycles and better logging so responders can see what happened in minutes, not days. Federal, state, and local agencies should share indicators fast and drill for “hours, not days” incidents. These moves defend civil society, protect small businesses, and back the rule of law while keeping government overreach in check.

Important Limits and How to Read the Numbers

The current evidence shows selected high-impact cases rather than a verified count of all artificial intelligence-caused incidents. Reports document capability and speed across espionage and crime, but they do not yet prove a sector-wide rate increase with full time-series data. That caveat does not blunt the practical message for defenders: assume faster intrusions, prepare playbooks for compressed timelines, and deploy controls that work even when an attacker learns and adapts between steps.

Bottom Line for Readers Who Want Action, Not Excuses

America faces foreign intelligence services and cybercriminals who now use artificial intelligence to move faster and hit more targets. The Federal Bureau of Investigation’s alert and Anthropic’s reports show real victims and real impact, from data loss to supply-chain risk. President Trump’s administration can continue pressing agencies and vendors to share threat data quickly, support critical infrastructure upgrades, and hold bad actors accountable. Families and business owners can back leaders who put security, privacy, and reliability first.

Sources:

msn.com, anthropic.com, www-cdn.anthropic.com, red.anthropic.com, therecord.media