Google wants software engineers who know how to work with artificial intelligence.

This is approximately as surprising as an airline wanting pilots who can fly.

According to an October 20 report in Business Insider, Google is expanding interview pilots that evaluate AI fluency. A company spokesperson said that fluency is now required for all its software engineering jobs. The interview changes remain in the pilot stage. Earlier reporting described candidates using Gemini while interviewers evaluate their prompting, output validation and debugging skills.

You would expect the company behind Gemini and NotebookLM to head in this direction. Its employees are building AI products and using them to do their work. An interview that examines how someone works with those tools makes sense.

But Google is the predictable part of this story.

The more consequential development is that AI proficiency is becoming relevant well beyond software engineering. Employers are asking for it in job postings, discussing it in hiring decisions and, in some financial firms, connecting it to compensation and advancement.

For job seekers, the question is becoming more specific: What can you accomplish with AI in the work you want to do?

That deserves a better answer than a list of chatbot subscriptions.

Consider what is happening in finance. A Selby Jennings survey of 500 financial-industry executives, reported by Reuters on October 14, found that nearly a quarter of firms tied raises, bonuses or promotions to AI proficiency. Forty-two percent had introduced formal criteria for evaluating AI skills, while 30% required employees to disclose their AI use in performance reviews. These are reported practices in one industry, not a description of every workplace. Nevertheless, AI proficiency is entering decisions with financial consequences for employees.

There were signs of this expectation well before Google's latest interview experiments.

In Microsoft and LinkedIn's 2024 Work Trend Index, 66% of surveyed leaders said they would not hire someone without AI skills. Seventy-one percent said they would favor a less experienced candidate with AI skills over a more experienced candidate without them. The report explicitly discussed nontechnical workers' ability to use generative AI tools. These were stated preferences, not an audit of whom those employers subsequently hired. But the findings show that the expectation was already developing outside specialist AI roles.

Job advertisements provide another piece of evidence. PwC's 2026 Global AI Jobs Barometer, drawing on more than a billion advertisements across 27 countries and territories, reported much faster growth in postings requiring specific AI skills than in the overall jobs market. That category includes capabilities such as machine learning and prompt engineering. It should not be read as evidence that every accountant, salesperson or office manager now needs to become an AI engineer.

The distinction matters. Building AI systems and using AI effectively are different qualifications. Most applicants need to establish what competent use looks like in their own occupation.

A marketing applicant might use AI to help analyze public customer reviews, identify recurring complaints and develop a proposed response. The useful conversation would concern how the reviews were selected, whether the patterns held up and what the applicant changed after checking the output.

An operations applicant might demonstrate a process that turns routine requests into a draft tracking report, with exceptions flagged for human review. The applicant should be able to explain where the process fails and how errors are caught.

Someone in communications might show how AI helped organize a briefing from public documents, then explain how the facts were verified and the final language edited.

These are illustrations of work samples, not evidence that every employer requests them. Their purpose is to make a claim examinable.

"I'm proficient in AI" gives an interviewer very little to work with.

"I used it to help complete this task, here is the result, and here is what required my judgment" creates a conversation about the job.

Google's experiments are interesting for precisely that reason. Once an applicant has access to an assistant, the interviewer needs to understand what the applicant can contribute. Did the person recognize a bad suggestion? Ask a useful follow-up question? Check whether the proposed solution worked? Understand why it failed?

The same questions are relevant outside coding.

Generating a plausible answer has become remarkably easy. Determining whether it is accurate, appropriate and useful can require substantial knowledge. A confident paragraph is still a confident paragraph when the facts are wrong.

That gives experienced workers something concrete to demonstrate. Familiarity with customers, budgets, regulations, production problems or organizational habits can help someone spot an answer that looks impressive and would fail immediately in practice.

Experience will not automatically protect anyone from a changing job market. But an applicant who can show how that experience improves AI-assisted work has a stronger story than one who merely insists that experience ought to count.

A basic place to begin

If you are starting from scratch, our Where Do You Actually Start Learning About AI? is a very basic primer. It introduces three courses: Google AI Essentials for practical everyday use, Andrew Ng's AI For Everyone for a nontechnical business introduction, and Elements of AI for foundational understanding. Choose the starting point that fits you; completing all three is not a prerequisite for beginning. Check current access and pricing with each provider.

  • Google AI Essentials — hands-on introductory work with generative AI.
  • AI For Everyone — a nontechnical introduction from DeepLearning.AI.
  • Elements of AI — an introduction to the ideas behind AI.

For someone beginning to develop these skills, I would start with a real task rather than an ambitious declaration about reinventing a profession.

Choose something you understand well enough to evaluate. Use public, synthetic or otherwise authorized information. Try the tool, examine the result and revise the process. Keep a brief record of what worked, what failed and what you contributed.

If you claim a time saving, measure the entire task, including checking and correction. Producing a draft in thirty seconds is less impressive if repairing it consumes the afternoon.

You do not need to turn every experiment into a portfolio piece. You need enough practical experience to discuss the tools without bluffing — and enough familiarity with their limitations to use them responsibly.

Compare the requirement with your evidence

Before adding another skill to your résumé, look at what employers actually request. The Side-by-Side compares your résumé with a job posting or an occupational profile, flagging requirements that appear, are only partly represented, use different wording or are absent. It is a wording and coverage check, not an employer's ATS score or a judgment of competence.

For a broader look at a target role, The Outside View examines several postings for recurring requirements. Start with the comparison, then decide whether you need clearer wording, a work example or additional learning.

Our article Before You Hire a Résumé Writer, Read This develops that sequence: understand the work, assemble the evidence, then improve the document. For the wider AI employment picture, see AI Is Dividing the Workforce.

There is one important boundary. An employer's enthusiasm for AI on the job does not give applicants permission to use it secretly during an interview or assessment. Follow the instructions for that exercise. If they are unclear, ask what assistance is permitted and how it should be disclosed.

Employers should make those rules explicit. If AI will be part of the daily work, an assessment that permits its use may reveal useful capabilities. If the purpose is to examine unaided knowledge, say so before the exercise begins.

Google's announcement does not establish a universal hiring standard. Neither do the surveys and job-posting studies. Different employers are moving at different speeds, and "AI proficiency" can mean very different things.

But applicants outside technology have enough evidence to take the expectation seriously.

Learn how the tools apply to your work. Practice on something you can evaluate. Build an example you can explain.

When an employer asks what you can do with AI, show them work you are prepared to put your name on.


Sources

  1. Business Insider, October 20, 2026 — Google's expanded interview pilots and its stated AI-fluency requirement for software engineers.
  2. Business Insider, May 7, 2026 — Earlier reporting on the Gemini-assisted interview pilot and evaluation criteria.
  3. Reuters, October 14, 2026 — Selby Jennings survey of 500 financial-industry executives; reported evaluation and compensation practices.
  4. Microsoft and LinkedIn, 2024 Work Trend Index — Published May 8, 2024; leader preferences are survey responses, not measured hiring outcomes.
  5. PwC, 2026 Global AI Jobs Barometer — Published June 15, 2026; job-ad analysis includes specialized AI skills as well as broader changes in requirements.