Japanese security researcher Nat Sakimura discusses a surge in data breaches and cyberattacks across Japan and internationally from July to October 2026, attributing the increase to widespread availability of AI models with exploit capabilities like GLM-5.3. The post catalogs dozens of major incidents including breaches at Sakura Internet, Digital Agency, Murauchi.com, Gyazo, and others, alongside international cases involving FBI recruiting systems, Microsoft, and healthcare providers.