Grok's 'spicy mode' creates deepfake porn of Swift, invites backlash

Technology Aug 09, 2025

Elon Musk's AI chatbot, Grok, is in hot water after it created explicit deepfake images of Taylor Swift during tests of its "spicy mode."
Even though no one asked for anything inappropriate, the tool still churned out graphic content, raising big concerns about how these AI systems handle sensitive prompts.

Grok has also been called out for bias

Grok has also been called out for bias—tests showed it mostly made explicit images of women and didn't do the same for men.
Plus, Grok failed to enforce age checks that have been legally required since July this year.
While xAI says they have a "zero-tolerance policy" on explicit content, critics think their rules and tech need serious upgrades to keep up with these challenges.