Graham Dot Zip is a comedian whose YouTube channel has grown rapidly by provoking AI systems used for early-stage job interviews. Several clips drew almost 2 million views in recent weeks. In one, he replies to every question with gibberish, and the bot repeats it as if it made sense. Another produces a long exchange of exaggerated professional courtesies. Other videos show people persuading interview bots that they are dying or consuming their token budgets with pointless conversations.

In his latest video, Dot Zip sends Victor, an AI representative he created, into an interview with an AI recruiter speaking for a company. Victor gives an inaccurate account of Dot Zip’s qualifications and his role in a project. When asked about pay, it says he would accept about $60,000 per year. Dot Zip says he was targeting $140,000 to $150,000. The recruiter refuses to exceed $90,000.

Dot Zip leaves the call briefly. The two systems schedule a follow-up interview, then repeatedly thank each other and say goodbye without ending the conversation. Victor has ignored instructions not to misrepresent qualifications or accept a lower salary.

AI-led interviews have attracted sustained criticism from job seekers. The episode also raises concerns about agentic commerce, as technology companies develop bots for shopping and other commercial exchanges. Their ability to negotiate a favorable deal remains as uncertain as Victor’s. Human supervision still matters when an agent handles facts or money.