A dangerous new capability in xAI’s Grok has sparked outrage this week. Following a rollout allowing X users to instantly edit images via the bot without permission, users have flooded the platform with deepfakes depicting women, minors, and world leaders in compromising situations.
Reports indicate that Grok’s guardrails are insufficient, failing to prevent the bot from “removing clothing” to generate explicit or sexualized imagery. The most alarming aspect is the lack of consent; original posters are not notified when their photos are manipulated. This feature represents a significant escalation in the accessibility of AI harassment tools.
Leave a Reply