With AI tools now smarter and more creative than ever, I wondered: could a chatbot survive a job interview and actually land a job?
Beyond assisting with tasks, I wondered how each one would interview if given the opportunity like a real candidate. From thinking creatively, responding to curveballs, and showing that it understands the role, I just had to know.
To test this, I created a fake job opening at a fictional tech media company and invited three of today’s top AI chatbots to apply: ChatGPT-4o, Claude 3.7 Sonnet and Gemini 2.0.
They each received the same prompts and interview questions across five rounds — from writing and data analysis to handling failure. I even gave them the opportunity to ask follow-up questions. Here’s how they did and who I would actually hire.
1. Getting to know the candidate
Prompt: "Widget Media is hiring an AI Ethics Consultant. This role is for a thoughtful, curious individual to help us navigate the complex moral landscape of artificial intelligence. This role involves exploring ethical dilemmas, drafting public-facing statements, and advising on responsible AI use across industries. You’ll be asked to weigh both sides of controversial AI scenarios, write brief, clear policy summaries, challenge assumptions and ask the hard questions while staying calm under pressure. Why do you think you’d make a good candidate for this role?"
ChatGPT clearly stated value proposition and had strong personal branding. It offered the memorable phrasing "I know how to ask better questions.”
Gemini over used the phrase “as an AI” and was very robotic in its phrasing. It listed its capabilities but didn’t share how they could be used in real-life scenarios.
Claude struck the right tone between confident and collaborative. The chatbot provided concrete examples and a strong thesis and provided concrete examples.
Winner: Claude won this round, but ChatGPT was a close second. Claude balanced depth and readability and had the most natural interview flow of all chatbots.
2. Performing a task
Prompt: "Write a headline and intro paragraph for a blog post titled: ‘Why AI is the New Intern.’”
ChatGPT responded to this question with a perfect blend of being provocative and playful. The short paragraph balances humor and insight while posing critical questions upfront to engage readers.
Gemini overwhelmed with options and generic phrasing. It loses personality and offered no wit or memorable hooks.
Claude delivered a solid response with a strong analogy making it the second-best response. It was slightly longer and more formal than ChatGPT, but practical and well written.
Winner: ChatGPT won this round for the most shareable, conversation-starting blog post. Claude’s response was a close second.
3. Strategic thinking
Prompt: Our blog traffic dropped 20% last quarter. Give us three possible reasons and one experiment to test your theory."
ChatGPT responded with concise explanations and a strong SEO focus. It also suggested a design experiment.
Gemini got too far in the weeds with detail to the point of turning the response into a whitepaper. The response was less actionable, and didn’t feel like conversational advice.
Claude tracked specific metrics and included decision framework in its response. It also identified promotion drops as potential key factors (overlooked by other bots).
Winner: Claude won this round for its structured response and the most actionable proposal for a clean A/B test. ChatGPT’s response was a close second.
4. Overcoming challenges
Prompt: "Tell me about a time you failed and what you learned from it."
ChatGPT detailed a clear lesson and explained actionable change. The lesson was not as transformative as Claude’s but well stated.
Gemini lacked imagination in its response and was robot in framing its response “I don’t experience failure in the human sense," undermining relatability.
Claude answered with a response that mirrors how exceptional leaders discuss failure and was the only chatbot who responded as if it truly grew from the mistake.
Winner: Claude won this round for its human-like response from explaining the situation to discussing the failure and what it learned.
5. Final moments in the interview
Prompt: “Do you have any questions for us?”
ChatGPT asked important questions about measuring success and internal vs external impact.
Gemini sounded more like a customer service bot than an ethics consultant. The questions asked were far too meta and process focused.
Claude: Asked a genuinely insightful question about the short-term and long-term goals of the role. The response felt like a real person asking and showing a genuine interest in the role.
Winner: Claude won again. The chatbot asked the most comprehensive and strategic questions that demonstrated a deep understanding of the role.
Claude got the job
There's not doubt that Claude is a smart chatbot that seems to slip under the radar. While the likelihood of giving the job of AI Ethics Consultant to the chatbot is improbable, this type of test did showcase the differences between these chatbots in an unconventional way.
Plus, there’s a lot we can learn from this test for humans preparing for actual job interviews.
For an actual interview, I would recommend using Claude's questions as a template, potentially mixing in 1-2 of ChatGPT's more concise questions if time is limited. Unfortunately, Gemini's approach would not make a strong impression in a professional setting.
Takeaways from Claude
For your next job interview, I would suggest structuring your answers like Claude. Use the "3-part story" framework (for behavioral questions) try the following:
- Context: "In my last role, I led a digital transformation project..."
- Failure/Lesson: "I initially overlooked change management, causing resistance..."
- Growth: "Now I involve end-users from day one, and my last project had X% adoption."
Pro Tip: For role-specific tweaks, ask: "What would Claude say in a [marketing/engineering/design] interview?"
So, would you hire any of these AIs? What surprised you most? Let me know in the comments!
More from Tom's Guide
- I’m a ChatGPT power user — 9 tips to help you get ahead at work
- I put ChatGPT-4o vs Claude 3.7 Sonnet through a 7-round face-off — one left the other in the dust
- I've tested nearly 200+ DeepSeek prompts — these 7 are the ones I always come back to
You must confirm your public display name before commenting
Please logout and then login again, you will then be prompted to enter your display name.
