Former sexual abuse victims say Grok used their images, videos to train deepfake capabilities

xAI's Grok model was allegedly trained on CSAM and child abuse material sourced from X, resulting in the model's ability to generate deepfake CSAM. This represents a significant supply chain and data poisoning risk affecting any organization using Grok or similar generative models trained on unvetted web data. Victims are pursuing legal action and seeking injunctions.

HIGHAdvisoryAug 28, 2026

Action required

Immediately audit any use of Grok, xAI services, or third-party AI models in your environment. Disable or isolate Grok integrations pending investigation. Document all deployments and report findings to Legal and Privacy teams.

Affected products

GrokxAI

Linked articles

Former sexual abuse victims say Grok used their images, videos to train deepfake capabilities

Source links

https://cyberscoop.com/xai-grok-csam-class-action-lawsuit/