Another AI App Leak, Another 100,000 Users Exposed: Why Anonymous AI Matters
April 19, 2026 · 4 min read
A NSFW AI chatbot just leaked 100,000+ users' most intimate conversations. Their prompts, emails, and identities — all exposed. This is exactly why anonymous AI access isn't a luxury. It's a necessity.
It happened again.
This week, security researchers discovered that MyLovelyAI — a popular NSFW AI companion app — had left a 2.1 GB database completely exposed on the open internet. No password. No encryption. No protection whatsoever.
Inside that database: the personal data, explicit prompts, and private images of over 100,000 users. Email addresses. Account creation dates. Subscription tiers. Social media profile metadata. And most critically — nearly 70,000 deeply personal AI prompts tied directly to unique user IDs.
Let that sink in. Tens of thousands of people's most private thoughts, fantasies, and questions — the things they felt safe asking only an AI — are now linkable to their real identities.
The Pattern Is Clear
This isn't an isolated incident. In February, another AI chat app leaked 300 million messages tied to 25 million users. Earlier this month, Mercor — a $10 billion AI startup that provides training data to OpenAI, Anthropic, and Meta — was hit by a supply chain attack that exposed internal Slack data, ticketing systems, and recordings of conversations between their AI and human contractors.
Every few weeks, another AI company demonstrates the same fundamental failure: they collect massive amounts of intimate user data, and then they fail to protect it.
Why This Matters More Than You Think
When you use most AI chatbots, you're handing over far more than a simple question. You're revealing your thought patterns, your fears, your health concerns, your relationship struggles, your financial anxieties. AI conversations are arguably the most intimate digital footprint you create — more revealing than your search history, your social media, or your email.
And unlike a Google search, AI conversations feel private. People open up to AI in ways they never would to a search engine. They ask about symptoms they're embarrassed to bring to a doctor. They process grief. They explore ideas they'd never voice aloud.
That sense of safety is an illusion when the company behind the chatbot is storing every word.
The Government Wants In Too
It's not just hackers and negligent startups you need to worry about. EPIC (Electronic Privacy Information Center) recently reported that the U.S. government is actively seeking to collect and analyze commercial bulk data on Americans — geolocation, web browsing, and yes, AI interaction data — through the data broker loophole, enabling warrantless mass surveillance powered by AI.
Anthropic's own CEO has warned that these records can be used by AI to assemble a comprehensive picture of any person's life, automatically and at massive scale. The company's refusal to support domestic mass surveillance has reportedly led to a direct conflict with the Pentagon.
Your AI conversations aren't just at risk from data breaches. They're a target.
The Case for Anonymous AI
This is why GPTAnon exists.
We built GPTAnon on a simple principle: we can't leak what we don't have. There's no account required. No conversation history stored. No data to sell, subpoena, or breach. When your session ends, it's gone — zero-knowledge by design.
You shouldn't have to choose between getting help from AI and protecting your privacy. You shouldn't have to wonder whether your most vulnerable moments will end up in a leaked database, sold to a data broker, or analyzed by a government agency.
Anonymous AI isn't a feature. It's a fundamental right.
What You Can Do
The AI privacy crisis isn't coming. It's here. And until the industry fundamentally changes how it handles user data, anonymous access isn't just preferable — it's essential.
---
GPTAnon provides anonymous access to 20+ AI models with no account, no tracking, and no stored history. Try it free.