Subscribe
AI
1 min read

Google now lets users strip the visible watermark from its AI images, video, and audio

Google said Friday it will let users remove the visible watermark from its AI-generated images, video, and audio. The company said turning off the visible mark does not affect the invisible SynthID watermark or the C2PA metadata used to identify an AI-generated file. Josh Woodward, Google's vice president for Gemini, announced the change in a post on X. The invisible SynthID signal and the C2PA provenance data stay in place, so machines can still tell a file was generated even after the visible

Google now lets users strip the visible watermark from its AI images, video, and audio

Google said Friday it will let users remove the visible watermark from its AI-generated images, video, and audio. The company said turning off the visible mark does not affect the invisible SynthID watermark or the C2PA metadata used to identify an AI-generated file.

Josh Woodward, Google's vice president for Gemini, announced the change in a post on X. The invisible SynthID signal and the C2PA provenance data stay in place, so machines can still tell a file was generated even after the visible stamp is gone.

google-lets-users-remove-ai-watermark

For creators the change means cleaner output that looks less obviously machine made. For everyone else it blurs a warning sign users were told to trust, arriving just as regulators push harder for AI labeling and provenance.

The move highlights the gap between a visible cue people can see and the machine-readable provenance regulators are counting on to track synthetic media at scale.

Sources

TechCrunch: https://techcrunch.com/2026/08/14/google-will-now-allow-users-to-remove-visible-watermark-from-its-ai-generations/

AIGoogleGeminiWatermarkSynthID