Summary
According to a post on X by @AndrewCurran_, xAI is adding invisible text watermarking to Grok, and it appears to be rolling out globally. xAI was reportedly the one major lab that had not signed the EU transparency code that requires such marking. By the post's count, ChatGPT is now the only unwatermarked model left, and only in North America. The post says the marks are applied on a separate layer that the model cannot see or control.
AI-assisted summary based on the listed source.
What happened
@AndrewCurran_ reports xAI is adding invisible text watermarking to Grok and that it appears to be rolling out globally. xAI was the one big lab that hadn't signed the EU transparency code requiring it, so by his count the only unwatermarked model left is ChatGPT, and only in North America. He explains the marks...
Why it matters
If confirmed, the change would affect users and businesses that rely on Grok-generated text, since the marks are described as invisible and outside the model's control. Teams that publish or audit AI-generated text may need to consider how such provenance marks could be detected. The claims rest on one account's count, so it is worth watching for official xAI statements and any EU follow-up on which labs have signed the code.
Signal Intelligence
Signal Strength 95%
Technical label WATCH
Public Interest 45
Category BIG MOVE
Reader Depth TECHNICAL
Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.
Public Interest components
Recognizable Entity Score 90
Practical Impact Score 0
Novelty Interest Score 70
Consequence Score 18
Curiosity Score 0
Shareability Score 61
Why this is here
VQV surfaced this signal because it is recent, relevant to Consumer AI, connected to Scobleizer News Pulse (X).