McDonald’s Idiotic AI Hiring System Just Leaked Personal Data About Millions of Job Applicants

When LLMS models become more integrated into the platforms that determine daily life, the main defects have begun to the program’s security capabilities to appear.

McDonald’s is among the list of the increasing companies that quickly rented LLM Chatbots in their recruitment systems, where the consequences are cursed. It is called chatbot Paradox.AI, which McDonald’s calls “virtual employment assistant”, as Olivia.

Olivia is more happy than happy with the help of applicants to find jobs near them through a fake conversation, with a picture of a human worker to make the whole matter more strange.

As Chatbot, Olivia is not great. It consumes job seekers through a maze of incomprehensible personality tests and examination questions, while completing the hallucinations that one can expect with LLM.

However, for an infiltrator who knows how to break llms, Olivia is a treasure awaiting detection.

As I mentioned for the first time before WirelessOlivia had some amazing security defects hiding under its human skin. Through the correct knowledge, the infiltrators can access chat nurses with 64 million applicants in McDonald, including personal details, such as full names, email addresses, phone numbers, addresses, work availability and crude chat data.

Cyber security researchers, Ian Carol Wissam Carrie, discovered the weakness in Olivia, by cyber security researchers, Ian Carole and Wissam Carrie, who managed to storm the back interface from the paradox. From there, infiltrators from White Hat managed to reach the “Test Restaurant Restaurant” of AI, giving them a glimpse of how everything works.

“It turned out that we became responsible for a test restaurant within the MCIRI system,” Carroll wrote on his blog. Since they are still confined to the Paradox.ai testing program, the husband decided to apply for one of the experimental publications to study the process.

By diving in the code behind the app, they quickly found a parameter indicating their application number, 64,185,742. When they tried to follow the application below their progress directly, they came face to face with the personal information of another job researcher, and revealed the whole world to see it.

“We realized this quickly [system] Carroll wrote: “We are allowed to reach every chat reaction that I applied for a job in McDonald’s.”

“We immediately started revealing this issue as soon as we realized the potential effect.” “Unfortunately, the disclosure contacts were not available to the public and we had to resort to sending an email to random people.”

“The Paradox.AI security page only says we do not worry about security!” He added.

Since then, the internal weakness has been corrected by Paradox, and the password of the official is no longer 123456. However, it is a clear view of the lack of responsibility for the adoption of early LLM.

More about piracy: Iranian infiltrators threaten the issuance of stolen emails from the Trump Interior Circle