> Examples include the central management of security keys and operating system updates; the scanning of content, like your e-mails, for advertising purposes;

> Examples include the central management of security keys and operating system updates; the scanning of content, like your e-mails, for advertising purposes; Now that Google, an advertising company, has given up scanning the contents of emails for advertising purposes: https://blog.google/products/gmail/g-suite-gains-traction-in-the-enterprise-g-suites-gmail-and-consumer-gmail-to-more-closely-align/ https://blog.google/products/gmail/g-suite-gains-traction-in... the necessity/reasonableness of this practice is highly questionable, but I suppose "for spam-filtering purposes" would be a clever alternative argument there. (Even with encrypted emails, though, there might be enough unencrypted metadata for accurate spam detection). Perhaps the real lesson for us here is the danger of centrally managed operating system updates. The fact that Rosenstein is already aware of this attack vector, and thinks so highly of it, suggests it could be a priority to address this. One technology that would be helpful here is Binary Transparency: https://wiki.mozilla.org/Security/Binary_Transparency https://wiki.mozilla.org/Security/Binary_Transparency so a user can at least be sure that they are being given the same updates as everyone else; and if researchers can audit the updates they can hopefully find "if userId = 123456" payloads. You'd have to trust the OS to not support "secret updates" which happen without your consent, but ideally this too would be addressed by an initial audit, or a TOFU policy.