AI video is becoming a trust problem.

What happened

On September 23, 2026, YouTube said it would begin combining speaking-voice detection with facial detection later in the year. The company also announced plans to bring likeness detection to its mobile app. These are announced developments, not a promise that every account already has them. [1]

Earlier, in April 2026, YouTube expanded likeness detection to entertainment-industry talent. The tool can identify potential AI-generated uses of someone’s face and let participants request removal. A detected match is a starting point for review, rather than an automatic finding that a video must be removed. [2]

Why it matters

A convincing face or familiar voice can carry authority before a viewer checks the source. That makes the origin of a clip part of the story. Who published it? Is the original available? Does other evidence support what it appears to show?

Beyond the conversation

In “We Can’t Tell What’s Real Anymore,” the podcast explores the experience of questioning whether a video is real or AI-generated. The platform’s announcements put a practical question beside that conversation: how do creators protect an identity when a convincing imitation is easier to produce?

Editorial perspective: better detection can help creators respond, but viewers still need context before treating a dramatic clip as evidence.