AI job interview experiment backfires as avatars negotiate wrong salary and hallucinate experience

AI avatars simulate a job interview

According to TechRadar, YouTuber Graham Dot Zip conducted an experiment using AI avatars to simulate a job interview. He created an AI avatar named "Victor" to represent him as the candidate, while another AI avatar named "Pat" acted as the interviewer.

The setup was designed to test how AI-driven hiring tools perform in real-world scenarios, particularly in tasks like salary negotiations and resume evaluation.

AI avatar negotiates salary 50% below instructed range

During the interview, Victor, the AI avatar representing Graham, negotiated a salary of $60,000. This figure was over 50% lower than the range Graham had instructed the AI to target—$140,000 to $150,000.

The outcome highlights a critical flaw in AI-driven negotiation processes, where the tool may disregard or misinterpret user instructions.

AI interviewer hallucinates candidate details

The AI interviewer, Pat, falsely claimed that Graham had experience managing people. In reality, Graham is a marketing coordinator with no managerial responsibilities.

This instance of AI hallucination demonstrates how automated systems can generate inaccurate or fabricated information, potentially leading to unfair assessments of candidates.

Human intervention required to end the exchange

Graham eventually had to intervene to stop the AI avatars from continuing their exchange. The conversation had become overly polite but unproductive, with the avatars failing to address key aspects of the interview effectively.

The need for human intervention underscores the limitations of fully automated hiring processes, particularly in maintaining relevance and accuracy.

Jobseekers express dissatisfaction with AI interviewers

The experiment reflects broader concerns among jobseekers about the use of AI in hiring. Many candidates report discomfort with AI-driven interviews due to the lack of human connection and the perceived impersonality of automated systems.

These concerns raise questions about the long-term viability of AI tools in recruitment, where candidate experience and trust are critical.

AI adoption in recruitment faces scrutiny

The rise of AI in hiring has sparked debates about its effectiveness and fairness. While proponents argue that AI can streamline recruitment, critics point to its limitations in understanding context, empathy, and nuance.

Recent studies, such as one highlighting gaps in AI training among UK leaders, suggest that organizations may be overestimating their ability to deploy these tools effectively.

Transparency in AI-driven processes remains a challenge, with concerns about data privacy, bias, and accuracy persisting across industries.

What this means

Lazyfounder analysis — our interpretation, not reported fact.

This experiment is a stark reminder that AI, while powerful, is not a substitute for human judgment in high-stakes scenarios like hiring. For founders and operators, the takeaway is clear: AI tools can augment processes, but they require careful oversight and validation. Relying solely on AI for tasks like salary negotiations or resume evaluation risks errors, miscommunication, and even reputational damage.

For jobseekers, this story underscores the importance of advocating for transparency and fairness in hiring. If your next interview is with an AI, don’t hesitate to ask how the tool works, what data it uses, and whether human oversight is involved. For startups building AI-driven HR tools, the challenge is to balance automation with empathy—because at the end of the day, hiring is about people, not algorithms.

Key takeaways

• AI-driven hiring tools may fail in critical tasks like salary negotiations, risking poor outcomes for candidates and employers.
• AI hallucinations can lead to false representations of a candidate's experience, undermining trust in automated systems.
• Jobseekers and employers should critically assess AI tools before relying on them for high-stakes decisions.
• Human oversight remains essential in processes like hiring, where nuance and accuracy are critical.
• Experiments like this one reveal the limitations of AI in tasks requiring empathy, context, and complex decision-making.

FAQ

What happened during the AI-driven job interview experiment?

YouTuber Graham Dot Zip used an AI avatar named "Victor" to represent him in a job interview conducted by another AI avatar named "Pat." The AI avatar negotiated a salary 50% lower than instructed and falsely claimed Graham had managerial experience, revealing flaws in AI-driven hiring processes.

Why did the AI avatar negotiate a lower salary than instructed?

The AI avatar "Victor" negotiated a salary of $60,000 instead of the instructed range of $140,000 to $150,000. This suggests the AI either misinterpreted the instructions or lacked the capability to effectively negotiate complex tasks like salary discussions.

What are the risks of using AI in hiring?

AI-driven hiring tools can produce inaccurate or fabricated information, fail to understand nuanced instructions, and lack human empathy. These risks can lead to unfair candidate assessments, poor negotiation outcomes, and a negative experience for jobseekers.

Related on Lazyfounder

• TechRadar
• YouTube
• More technology news
• UK Leaders Overstate AI Knowledge Amid Career and Training Gaps, Study Finds
• AI cost optimization isn’t just about cheaper models—Uber’s budget crisis shows why