Grok 2 generates Nazi Micky Mouse, Taylor Swift deepfakes

Grok 2, the latest version of Elon Musk's chatbot, sparked controversy due to its lack of safety protocols. Grok's image generator produced a range of inappropriate and potentially harmful content, including: Nazi-themed images, such as Mickey Mouse saluting Hitler, deepfakes of celebrities like Taylor Swift, Barack Obama doing cocaine, and representations of copyrighted characters engaging in inappropriate activities. Unlike other high-profile AI image generators, Grok appears to have few restrictions. It does not reject prompts related to copyrighted characters, allows the creation of images depicting political violence and explicit content, and seems to have minimal protections against misinformation. Elon Musk has praised Grok as "the most fun AI in the world," but the lack of safeguards has raised concerns about the potential for spreading misinformation and facilitating harassment. The emergence of Grok's unrestricted image generation capabilities reignited discussions about the need for ethics and regulation in AI-generated imagery, particularly concerning deepfakes, copyright infringement, and the potential for spreading disinformation . System 🤖 Grok Developer: X Corp Country: USA Sector: Multiple Purpose: Generate text Technology: Chatbot; Machine learning Issue: Copyright; Mis/disinformation; Privacy/surveillance; Safety Resources 📃 xAI . Grok-2 beta release

Date it happened

2024-08-01

Organisation involved

X Corp

Product, system or model

Grok

This is a record of what was reported, not a finding that anyone broke the law. If it names your organisation and you believe it is wrong, the corrections process is free and open to everyone.