AI Questions Now Appearing in Job Interviews, With Good Answers

Kenyan employers across banking, marketing, operations, and administrative roles are now asking AI-related questions in interviews. These are not deeply technical questions about machine learning. They are practical questions about how you use AI tools, whether you understand their limitations, and if you can apply them to real work. Here are the most common ones, with answers that.

Why AI questions are showing up in non-technical interviews

Two years ago, AI questions appeared only in tech interviews. That has changed. Hiring managers at Kenyan banks, NGOs, retail companies, and service firms now include AI questions because they want to know if candidates can work productively with these tools.

The questions are not designed to test whether you can build a neural network. They test whether you have practical experience, whether you understand what AI tools can and cannot do, and whether you can think critically about when and how to use them.

If you are preparing for interviews, treating AI questions as "bonus" questions is a mistake. For many roles, they are now standard.

Question 1: "How do you currently use AI tools in your work?"

What the interviewer wants to know

They want to understand whether you have actual hands-on experience, not just awareness that AI exists. They are listening for specific tools, specific tasks, and specific results.

A good answer

"I use ChatGPT regularly for three things: drafting first versions of client emails, which I then edit for tone and accuracy; summarizing long research documents so I can quickly identify relevant sections; and generating spreadsheet formulas when I need something beyond basic functions. I also use Make to automate our weekly report compilation. The report used to take about two hours of manual copy-pasting from three different sources into a Google Sheet. The automation runs every Monday morning and has the report ready before the team meeting."

What makes this answer work

It names specific tools, specific tasks, and a specific result. It shows the candidate edits AI output rather than blindly accepting it. It demonstrates both AI tool usage and automation skills.

A weak answer

"I use ChatGPT sometimes to help with various tasks at work." This tells the interviewer almost nothing.

Question 2: "What are the limitations of AI tools you have noticed?"

What the interviewer wants to know

This is a maturity test. Anyone can be enthusiastic about AI. The interviewer wants to know if you understand the boundaries, because someone who does not understand limitations will make costly mistakes.

A good answer

"The biggest limitation I have seen is accuracy. AI tools generate confident-sounding text even when the content is wrong. I had ChatGPT produce a competitor analysis that included a company's 'revenue figures' that were entirely fabricated. It presented them as facts. Since then, I verify any specific claims, numbers, or data points that AI generates before including them in any work product. I also found that AI tools struggle with specifically Kenyan context. They might suggest marketing strategies that work in the US but are irrelevant here, or reference regulations that do not apply in Kenya. Human judgment about local context is still essential."

What makes this answer work

It shows personal experience with a specific failure. It describes how the candidate adapted their process. It demonstrates critical thinking about AI output.

Question 3: "How would you use AI to improve a process in this role?"

What the interviewer wants to know

They want to see if you can look at their business problems and identify where AI is genuinely useful, not just insert AI into every answer because it sounds modern.

A good answer

"Based on what you have described about the role, I see two areas where AI could help immediately. First, the weekly client reporting you mentioned. If the data sits in spreadsheets and a CRM, I could set up an automation that pulls the numbers, uses AI to generate a plain-language summary paragraph, and sends a draft report to the team lead for review. Second, the customer inquiry handling. If you receive similar questions repeatedly, an AI-assisted template system could draft initial responses that the team reviews and personalizes before sending. I would not use AI for the client strategy discussions you described, though. Those require understanding each client's specific business situation, which AI cannot do well from data alone."

What makes this answer work

It identifies specific, practical applications. It connects to what the interviewer described about the actual role. It also identifies where AI should not be used, which shows judgment.

Question 4: "Have you automated any part of your workflow? Tell us about it."

What the interviewer wants to know

They are testing whether you take initiative to improve processes, not just follow existing ones. Automation experience signals that you think about efficiency and can implement solutions.

A good answer

"Yes. At my previous role, every time we received a new partnership inquiry through our website form, someone had to copy the details into a spreadsheet, send a confirmation email to the inquirer, and notify our partnerships manager. It took about ten minutes per inquiry and sometimes got delayed on busy days. I set up a Make automation that does all three steps instantly when the form is submitted. The form response goes straight to a Google Sheet, a confirmation email goes out within a minute, and the partnerships manager gets a Slack notification with the key details. It has been running for about five months without issues."

What makes this answer work

It describes a real problem with specifics. It explains the solution and the tools. It shows the result and its reliability.

Question 5: "How do you verify that AI output is accurate?"

What the interviewer wants to know

This question separates people who use AI thoughtfully from those who copy-paste whatever comes out. Given that AI tools confidently produce incorrect information, verification is a professional skill.

A good answer

"My process depends on the type of output. For factual claims or data, I cross-reference with original sources. If ChatGPT says a company launched a particular product, I check the company's website or a news article. For written content, I read every word and edit for accuracy, tone, and fit. I do not send AI-drafted text without reviewing it. For code or formulas, I test them with sample data to make sure they produce the expected results before applying them to real data. The general rule I follow: the higher the stakes of the output, the more verification I apply."

What makes this answer work

It describes a systematic approach. It acknowledges different levels of verification for different uses. It shows professional responsibility.

Question 6: "Are you comfortable learning new AI tools as they evolve?"

What the interviewer wants to know

AI tools change rapidly. The interviewer wants to know if you will keep up or if your current knowledge will become outdated.

A good answer

"Yes. I have already gone through several tool transitions. I started with basic ChatGPT usage, then learned prompt engineering techniques to get better results, then added Claude as a second option for different types of tasks, then started using Make for automation. Each new tool took a few days of focused practice to become comfortable with. My approach is to dedicate about thirty minutes each week to exploring updates in the tools I use and trying one new feature or tool each month. I find that consistent small investments in learning prevent the need for large catch-up sessions."

What makes this answer work

It shows a track record of learning. It describes a sustainable learning habit. It is realistic rather than vague promises about being a "fast learner."

Preparing for AI interview questions

Build real experience before the interview

The answers above work because they draw on real experience. If you have not used AI tools meaningfully, start now. Even a few weeks of deliberate practice gives you genuine examples to discuss.

Prepare three to four specific examples

Before any interview, prepare stories about: one time AI helped you complete a task faster or better, one time AI produced an error and how you handled it, one automation or process improvement you implemented, and one area where you chose not to use AI and why.

Practice honest answers

Interviewers can tell when you are fabricating experience. If you have limited AI experience, say so honestly and describe what you are learning. "I have been learning automation skills through the AI and Automation for Beginners course and have built three practice automations so far. I am ready to apply these skills in a professional setting." This is more credible than exaggerated claims.

Understand the role's context

AI questions are better answered when you connect them to the specific role. Research the company beforehand. Think about what processes they likely do manually and where AI could help. Tailor your answers to their context, not to generic AI use cases.

FAQ

What if I have no AI experience and the interview asks about it?

Be honest, then redirect to your learning mindset and transferable skills. "I have not used AI tools extensively in my previous role, but I have started learning through online courses and practice projects. I am currently comfortable with ChatGPT for drafting and research, and I am building skills in workflow automation. My experience in [your field] means I can identify where these tools add value quickly." Honesty about your current level, combined with evidence of initiative, is better than bluffing.

Do Kenyan employers expect the same AI skills as international ones?

The core expectations are similar (using AI tools for productivity, understanding limitations), but Kenyan employers may place more emphasis on local context: M-Pesa integrations, WhatsApp-based customer communication, and understanding of the local business environment. International employers (for remote roles) may expect familiarity with a broader range of tools and more experience with platforms like Slack, Notion, and HubSpot.

Should I mention AI skills even if the job posting does not specifically ask for them?

Yes, if you can connect them naturally to the role's responsibilities. Most job postings have not caught up with how much AI is now used in everyday work. Mentioning relevant AI skills shows initiative and current awareness. Just make sure the connection to the role is clear, not forced.