In response to the EU AI Act, text generated by Grok models will carry an invisible watermark
Starting November 7, SpaceXAI will add an invisible watermark to text generated by Grok models; detectors can identify this signal to determine whether text carries a watermark invisible to the naked eye. Officials say the move is in response to the EU Artificial Intelligence Act. The watermark has been enabled across SpaceXAI APIs in all regions worldwide, covering the reasoning, non-reasoning, and multi-agent versions of grok-4.7, grok-4.6, grok-4.5, grok-4.3, and grok-4.20, and team administrators can disable the watermark for an individual model. The watermark is not a hidden character or metadata, but is embedded in the model's token selection process when generating text.
Why read this
AI score 59
The original text explains the range of models covered by the watermark and the opt-out option, offering insight into how the EU AI Act's requirement to label generated text is being implemented in practice.
Topics
Body · AI translation
Responding to the EU AI Act, Text Generated by Grok Models Will Add Invisible Watermarks
2026-11-01 14:11 IT Home - Luyuan
IT Home, November 1 news: SpaceXAI will add invisible watermarks to text generated by Grok models starting November 7. Detectors identify this signal to determine whether a piece of text carries a watermark invisible to the naked eye.
The company stated that this adjustment is in response to the requirements of the EU Artificial Intelligence Act. Watermarks have been enabled for SpaceXAI APIs in all regions worldwide, covering the grok-4.7, grok-4.6, grok-4.5, grok-4.3, grok-4.20 reasoning, grok-4.20 non-reasoning, and grok-4.20 multi-agent models. Team administrators can choose to disable watermarks individually for a given model.
IT Home notes that this hidden watermark is not implemented through hidden characters or metadata, but is instead integrated into the process of selecting Tokens when the model generates text.
AI governance observer Andrew Curran (IT Home note: Andrew Curran) said in response that currently the only model that has not added invisible watermarks to text is ChatGPT in North America.