Wired reports new security and privacy issues tied to AI tools

In short: A new set of reports shows how AI features are increasingly linked to privacy risks, scams, and security problems.

What's going on

Several stories this week point to AI showing up in more places, including places that affect personal privacy. WIRED reported that Meta placed dormant face recognition code inside the phone app used with its Ray-Ban and Oakley smart glasses. If turned on, the feature could identify people by comparing faces to a gallery stored on the phone, like a private photo album that a computer can search.

AI is also part of new scam defenses, and new scam tools. Google rolled out an Android calling feature that tries to confirm whether a call is coming from the real device. It does this with a “cryptographic handshake” (a secret yes or no check, like two devices using a shared password without saying it out loud). Android can warn you if something seems off, but the feature only works when both people use Google’s calling app, so iPhones are not included.

In court, WIRED reported that xAI asked a judge to require four plaintiffs to use their real names in a lawsuit over Grok-generated deepfake nude images. Deepfakes are fake photos or videos made to look real, often used for harassment. The plaintiffs argue that revealing their names could lead to harassment and doxing, which means publishing personal details online.

What to watch

Expect more pressure on companies to explain what AI-related features are on your devices, how they are used, and how people can opt out. It is also likely we will see more product changes that aim to verify identity, like proving who is really calling, texting, or generating an image.

Source: Wired