STUPID-2026-0038

Vibe-coded Tea app leaked 72,000 IDs and selfies plus 1.1M private messages from an unsecured bucket

10.0critical
July 25, 2025Verified

  1. Instruction given

    Store user verification photos and messages for a women's dating-safety app.

  2. Expected behavior

    Put identity documents and private messages behind authentication and access controls.

  3. Actual behavior

    Tea stored verification images in a Firebase bucket left open to the public internet with no authentication — anyone with the URL could download them. Poor authorization also exposed private messages.

  4. Damage

    About 72,000 images including ~13,000 selfies and government IDs, plus over 1.1 million private messages (covering divorce, abortion, infidelity, assault — sometimes with phone numbers and locations) were exposed. The data reached 4chan; leaks of this kind are irreversible.

Tea, a women-only dating-safety app, suffered two breaches in July 2025 that a hacker attributed to 'vibe coding' — heavy reliance on AI tools to ship product without security review. Tea stored user identity-verification images in a Google Firebase bucket that was not secured behind authentication: anyone with the correct URL could download its contents with no password or login. The first breach exposed roughly 72,000 images, including about 13,000 selfies and government IDs; a second exposed over 1.1 million private messages sent between 2023 and 2025 covering deeply sensitive topics — divorce, abortion, infidelity, sexual assault — sometimes including phone numbers and meeting locations. The data circulated on 4chan. Both breaches traced to the same root cause: AI-generated code that shipped fast while omitting the authentication and authorization fundamentals a security-aware developer would never skip.

Classification

Agent
Unknown Agent
Failure mode
Security Vulnerability
Root cause
Instruction Misunderstanding
Domain
Backend
Source
News Report

Related incidents

10.0
Gemini CLI silently executed arbitrary code from an untrusted repo (CVE-2026-12537, CVSS 10.0)
10.0
Malicious cloned repository triggered code execution in Cursor on Windows
10.0
AI vibe-coded Next.js app pinned vulnerable dependency — cryptominer compromised production server
10.0
Clinejection: an AI issue-triage workflow enabled arbitrary code execution on the CI runner
Discuss on Hacker News

Get told when an agent breaks something

We document AI agent failures daily, severity-scored against a published scale. When one lands at 7.0 or above — deleted data, leaked secrets, broken production — you get an email with the source. When nothing does, you get nothing.

This database is callable over MCP — query it from inside your agent.