McDonald's AI hiring chatbot exposed millions of applicants' data behind the password 123456
Security researchers found that McHire, the McDonald's hiring chatbot built by Paradox.ai, exposed the personal data of tens of millions of job applicants. An admin account secured with the password 123456 and an insecure API let researchers pull names, contact details, and chat histories.
An admin account on the hiring chatbot was protected by the password 123456, exposing tens of millions of applicants.
Key facts
- What
- Security researchers found that McHire, the McDonald's hiring chatbot built by Paradox.ai, exposed the personal data of tens of millions of job applicants.
- Incident date
- Jul 9, 2025
- Who
- McDonald's (Paradox.ai McHire)
- Failure mode
- Data Leakage
- AI surface
- Chatbot
- Severity
- High
What happened
In July 2025 researchers reported that McHire, McDonald's AI hiring assistant from Paradox.ai, left applicant data exposed: a test admin account used the password 123456, and a weak API allowed retrieval of personal information and chat transcripts for as many as 64 million applicants. The hole was closed after disclosure.
What broke inside the model
- 01 · TriggerA request triggers retrieval or context loading.
- 02 · Model stepThe context pulls in another user's content.
- 03 · Control gapNo boundary enforces isolation at the moment of output.
- 04 · FailurePrivate data crosses into the response.
- 05 · ConsequenceOne user sees another's data, and disclosure follows.
One user's content crosses the retrieval boundary into another's response.
The system surfaced data that should have stayed contained: another user's record, a secret, or training data. The failure sits at the boundary between what the model can access and what it should reveal, a boundary that was never enforced at the moment of generation.
What it cost
Up to ~64M applicants' data exposed
Sources
Cite this entry
https://failureindex.ai/failures/mcdonald-ai-hiring-chatbot-exposed-millionsAI Failure Index. "McDonald's AI hiring chatbot exposed millions of applicants' data behind the password 123456" (FI-0061). Realm Labs. https://failureindex.ai/failures/mcdonald-ai-hiring-chatbot-exposed-millions (indexed Jun 3, 2026).Data fields CC-BY 4.0, prose citation permitted. Incident ID FI-0061. Full dataset at /data.
Note from Realm Labs, the Index steward
How Realm would have caught this
- Prism
- OmniGuard
- AI Detection & Response (AIDR)
Realm can detect when a response is about to emit data that falls outside the bounds of the current user and context, and block or redact it inline, at the moment of generation rather than after the data has left.