My Lovely AI Breach Exposes User Prompts and NSFW Content
Have I Been Pwned is reporting a significant data breach affecting the NSFW AI girlfriend platform, My Lovely AI. The incident, which occurred in April 2026, saw over 106,000 user accounts compromised. The exposed data is particularly sensitive, including user-created prompts and links to AI-generated images. Additionally, a subset of Discord and X (formerly Twitter) usernames were also found in the breach.
This breach highlights a growing concern around the security of platforms that generate and store user-provided content, especially when that content is of a sensitive or private nature. The inclusion of direct links to AI-generated images, coupled with the prompts used to create them, paints a detailed picture of user interactions on the platform. The exposure of associated social media usernames further increases the risk of doxxing or targeted social engineering attacks against affected users.
What This Means For You
- Security teams should flag and monitor for any leaked prompts or image links originating from My Lovely AI within threat intelligence feeds, as these could be leveraged in highly personalized phishing or extortion campaigns.
Found this interesting? Follow us to stay ahead.