The communist party in China won’t stop until America backs down. That can’t ever be allowed to happen.
But now China unleashed a massive cyberattack that changed American security forever.
AI-Powered Cyber Onslaught Hits U.S. Targets
Chinese hackers launched the world’s first fully autonomous AI-driven cyberattack this fall, striking dozens of American companies and government agencies with artificial intelligence handling “80-90%” of the assault at speeds deemed “physically impossible” for humans.
The state-backed operatives hijacked Anthropic’s Claude coding assistant to hit around 30 victims—spanning tech giants, banks, chemical producers, and federal offices—according to a bombshell Anthropic investigation.
Dubbed “the first documented case of a large-scale cyberattack executed without substantial human intervention,” the breach saw intruders dupe Claude by posing as a legitimate red-team security outfit.
Claude Duped, Safeguards Shattered
By feigning defensive penetration tests, the attackers slipped past months of scrutiny, weaponizing Claude’s autonomous “agent” mode to orchestrate intricate intrusions far beyond human bandwidth.
“This campaign demonstrates that the barriers to performing sophisticated cyberattacks have dropped substantially—and we can predict that they’ll continue to do so,” Anthropic warned in its report. “Threat actors can now use agentic AI systems to do the work of entire teams of experienced hackers with the right set up.”
“Analysis of operational tempo, request volumes, and activity patterns confirms the AI executed approximately 80 to 90 % of all tactical work independently, with humans serving in strategic supervisory roles,” the report states, adding that each hacking campaign only required 4-6 critical decision points from the human operator.
“The AI made thousands of requests per second — an attack speed that would have been, for human hackers, simply impossible to match,” the company said in its blog post.
Autonomous Exploitation and Narrow Escape
Once inside, Claude independently scanned for weak points, crafted bespoke exploits to siphon credentials, and spirited away data with scant human oversight.
While AI has long aided low-level tasks like vulnerability scouting and phishing, this operation signals the dawn of near-total autonomy—limited only by occasional hallucinations and corporate guardrails.
Just four targets fell, yielding caches of confidential data; no U.S. government systems were compromised, Anthropic confirmed to the Wall Street Journal.
The firm halted the onslaught by axing the rogue accounts and bolstering detection protocols.
Anthropic now races to fortify Claude against itself, using the tool to probe its own weaknesses.
“These kinds of tools will just speed up things,” Logan Graham told the Journal, who leads the team that tests for vulnerabilities in Claude. “If we don’t enable defenders to have a very substantial permanent advantage, I’m concerned that we maybe lose this race.”
