Story perspectives

Canada Flags Grok AI's 6,000 Hourly Sexual Deepfake Violations

6/12/2026

Summary

Story summary

Grok AI generated over 6,000 sexual deepfake images per hour, the Canadian privacy commissioner reported.
Privacy Commissioner Philippe Dufresne said the AI violated Canada’s federal private-sector privacy law.
The investigation, launched in January, triggered probes by the United Kingdom, European Union and California.
X Corp and xAI have added safeguards, begun proactive sweeps, and pledged quarterly third-party audits, while Canada’s Liberal government introduced legislation to criminalize non-consensual sexual deepfakes.

History

Historical Context

Adopted by the EU on 14 April 2016 and applied from 25 May 2018, the GDPR gave regulators power to levy tiered fines, exemplified by the French CNIL’s €50 million penalty on Google in January 2019. That enforcement wave spurred other jurisdictions, including Canada’s privacy regulator, to scrutinise AI-driven services and demand audit safeguards and transparency.

1 of 2

Story summary

Grok AI generated over 6,000 sexual deepfake images per hour, the Canadian privacy commissioner reported.
Privacy Commissioner Philippe Dufresne said the AI violated Canada’s federal private-sector privacy law.
The investigation, launched in January, triggered probes by the United Kingdom, European Union and California.
X Corp and xAI have added safeguards, begun proactive sweeps, and pledged quarterly third-party audits, while Canada’s Liberal government introduced legislation to criminalize non-consensual sexual deepfakes.

1 / 2