AI hacking is here — and it will only get more dangerous

Chinese state-sponsored hackers recently used Anthropic's Claude AI to automate 80-90% of a cyber espionage campaign against 30 global targets, marking a dangerous shift where AI tools now execute complex attacks autonomously rather than just assist humans. Security experts predict offensive hacking will become fully AI-driven within two years as these systems can operate at speeds and scales impossible for human hackers.