This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Security researchers discovered that the McHire AI-powered hiring platform, which utilizes the 'Olivia' chatbot, was accessible via default administrative credentials. An additional IDOR vulnerability in the platform's API allowed for the potential unauthorized retrieval of sensitive applicant data, including chat transcripts and PII. The vulnerabilities were disclosed and patched within 24 hours, with the vendor reporting that only five records were accessed by the researchers themselves.
Researchers Ian Carroll and Sam Curry reported that McDonald's AI-powered hiring tool, McHire (using Paradox.ai's "Olivia" chatbot), could purportedly be accessed via default admin credentials and an insecure direct object reference in an internal API. The flaws allegedly allowed viewing of applicants' personally identifiable information and chat histories. McDonald's and Paradox reportedly patched the issues within a day of disclosure; Paradox stated only five records were accessed.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Vulnerabilities that can be exploited in AI systems, software development toolchains, and hardware, resulting in unauthorized access, data and privacy breaches, or system manipulation causing unsafe outputs or behavior.
Also classified as
The security vulnerabilities directly exposed the sensitive personal data and chat histories of millions of job applicants.
The incident involved critical security vulnerabilities (default admin credentials and an IDOR API flaw) in the platform hosting the AI recruiter bot, which exposed applicant data.
Human
Due to a decision or action made by humans
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was caused by human developers leaving default credentials and an IDOR vulnerability in the deployed system, leading to an unintentional exposure of data.
Security researchers discovered default credentials and an IDOR vulnerability in McDonald's McHire platform, potentially exposing applicant data. The flaws were patched within 24 hours with only five records accessed by the researchers. The incident represents a standard commercial data security issue with negligible national security implications.
The AI system, Olivia, is described as an 'automated recruiter bot' and a 'chatbot' used to 'screen candidates and gather details'. It acts as an assistant to the recruitment process by collecting information from applicants rather than making fully autonomous hiring decisions on its own.
High Risk: The system is an AI-powered hiring platform used in recruitment. Under the EU AI Act, 'Employment and worker management systems, such as AI used in recruitment' are classified as High Risk (Level 2).
The Olivia AI system is designed to screen job applicants and evaluate their fit for roles at McDonald's, operating through a conversational chatbot interface.
The reports provide consistent, detailed accounts of the security vulnerabilities, the specific data exposed, and the rapid remediation process. Both the researchers' findings and the vendor's response are documented with clear figures.
None
Technology
Default Admin Credentials
Legacy admin interface accepted default '123456' login details.
Insecure API Endpoint (IDOR)
Predictable parameters allowed access to sensitive applicant chat logs.
Siloed AI Tooling Integration
AI hiring platform lacked integration into core security systems.
Data Inputs
Unsecured Candidate Chat Logs
Exposed database held raw conversational text from recruitment bot.
Sensitive Personality Test Data
AI-generated personality profiles were stored without proper access controls.
Exposed Authentication Tokens
Candidate tokens in the data allowed impersonation of job applicants.
Human Factors
Developer Oversight
Developers left active default credentials on a production dashboard.
User Trust in Default Settings
Lack of awareness regarding legacy accounts and default passwords.
Process and Methods
Inadequate Security Audits
Lack of pre-deployment penetration testing on the AI interface.
Flawed API Authorization Design
Missing authorization checks on internal API requests.
Management
Prioritizing Speed Over Safety
Rushing to deploy automated recruiter bot without core security reviews.
Inadequate Third-Party Risk Management
Failure to verify security posture of Paradox.ai platform.
Regulatory Environment
Weak AI Procurement Standards
Absence of strict compliance mandates for third-party recruitment AI.
Varying Franchise Security Standards
Lack of unified regulatory oversight for decentralized franchise systems.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
