China's AI Hackers Just Got Scary Good and Wall Street Is Freaking Out Over This New Zero-Day Machine

This is your Digital Dragon Watch: Weekly China Cyber Alert podcast. Hey listeners, Alexandra Reeves here with Digital Dragon Watch, your weekly China cyber alert. Over the past seven days ending April 12, 2026, China's hackers unleashed an AI upgrade that's got Wall Street sweating and the US government scrambling. Picture this: elite Chinese state-sponsored groups, like those tracked by the National Cyber Security Centre, just leveled up with AI-assisted attacks ripping through public-facing apps. IBM reports a whopping 44 percent spike in exploits this year, fueled by tools that automate vulnerability hunting, craft personalized phishing in any language, and chain exploits into full campaigns. Red Canary says adversaries are leaning on large language models for 80 to 90 percent of their espionage ops—think reconnaissance and malware that evolves to dodge detection. Trend Micro dubs it the "AI-fication of cyberthreats," and it's hitting hard in telecoms, finance, and critical infrastructure. Targeted sectors? Cloud and SaaS setups are bleeding data from misconfigurations—publicly exposed storage, leaky APIs, over-privileged accounts. Tata Communications warns India's facing a 800,000 cyber pro shortage, but the ripple hits global players too, with hybrid workforces and remote ops erasing old network perimeters. Identity access management flops are now the top breach trigger, per their analysis. Enter the Dragon's latest twist: Anthropic's Mythos Preview, announced April 7. This beast nailed a 72.4 percent success rate in spitting out working zero-day exploits—leaps ahead of prior models. They gated it behind Project Glasswing, handing previews only to 40-50 critical infra giants like Amazon Web Services, Apple, Cisco, CrowdStrike, Google, JPMorgan Chase, Microsoft, NVIDIA, and Palo Alto Networks. No public release, because offensive AI power now laps defenses. US response? White House heavyweights are all in. National Cyber Director Sean Cairncross is rallying agencies to plug critical infra holes, beef up government systems against AI hacks, and sync with the Office of Science and Technology Policy and National Security Council. Wall Street Journal sources spill that interagency calls looped in Vice President Vance, Treasury Secretary Bessent, and execs from Anthropic, OpenAI, Microsoft, Google, CrowdStrike, and Palo Alto. They're prepping for Mythos's eventual drop to thwart attacks. Expert recs? Ditch VPNs and IP firewalls for zero-trust architectures—verify every access, everywhere. Tata pushes skilling programs to close talent gaps, weave AI into defenses, and hunt AI-driven threats. Secure those cloud misconfigs, lock down identities, and assume breach. China's 15th Five-Year Plan amps their cyber strategy, per This Week in 4n6, eyeing supply chains like Nebulock Hunt Mode. NCSC flags APT28 router exploits for DNS hijacking—watch your edges. Stay vigilant, listeners—patch fast, train up, go zero-trust. Thanks for tuning in to Digi This content was created in partnership and with the help of Artificial Intelligence AI.