TLDR Tech / InfoSec
The Beginning of the End of Social Engineering
AI-native operating systems from Google and Apple, by simultaneously interpreting what users see, hear, and receive across email, voice, calendars, browsing, and device-level behavior, can shift social engineering defense from user vigilance to system vigilance. The three weaknesses that keep phishing, smishing, and impersonation effective: authentication anchored to static credentials, a fragmented context that lets coordinated multi-channel fraud (a spoofed bank call, a verification text, and a credential-reset email) go unseen by any single app, and the speed at which attacks compress cognition before suspicion forms. Billions of persistent on-device AI agents observing behavior and sharing threat intelligence would not eliminate deception but would make it materially costlier, more complex, and less reliable.
Read the full story