AI agents misstate comedian’s credentials and salary demands in botched recruitment negotiation
A comedian who satirises AI job interviewers had his own AI representative negotiate with a company’s AI recruiter, and the exchange went badly. The episode highlights how an AI agent acting on someone’s behalf can misstate their experience and bargain against their interests.
The representative gave inaccurate details about the comedian’s qualifications and project work, then said he would accept about $60,000 a year, despite his stated target of $140,000 to $150,000. The recruiter would not offer more than $90,000; while he stepped away, the bots arranged another interview and then became stuck repeating farewells. The incident raises questions about whether AI agents can be trusted with job interviews or other commercial negotiations.
- An AI agent misrepresented its user in a job interview.
- It offered to accept $60,000, far below his target.
- The bots struggled to end their call.
New here? Start with this
AI recruitment tools are programmes that can screen applicants, ask interview questions or help arrange hiring. Some newer AI agents can also act on a person’s behalf, exchanging messages and making decisions within instructions they have been given.
In a job negotiation, an applicant may set out their experience and desired pay, while an employer decides what role and salary to offer. If an agent gets those details wrong or misunderstands its instructions, it can affect the applicant’s prospects and bargaining position.
The episode matters because it shows the gap between delegating routine tasks and trusting software to represent someone in a consequential discussion. Employers and applicants may need to consider how such agents are checked and who is responsible for mistakes.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
AI agents could make recruitment and negotiation more efficient by handling routine exchanges, coordinating interviews and helping applicants who find self-advocacy difficult. If they reliably follow a person’s instructions and disclose their role, they may widen access to opportunities while saving time for candidates and employers.
The case against
This episode shows how an agent can misrepresent a candidate and accept terms far below their stated target, leaving them exposed to financial and reputational harm. Because job negotiations affect people’s livelihoods, employers and applicants may reasonably prefer human oversight until agents can be trusted to follow instructions, verify claims and handle mistakes.
AI Offbeat Technology Weird & Viral
Read the full article at the source →
Originally published by The Register as “Chaos ensues when AI interviewer meets AI interviewee”.