Hiring managers have stopped taking AI claims on a résumé at face value, and they have a reasonable case. AI now writes a growing share of résumés, so the document carries less signal at exactly the moment the skill matters most.
Across sixteen interviews with senior leaders conducted by Clarity Recruitment between April and June 2026, the screen is moving in one direction: show me.
What changed in hiring
Two things happened at once.
The scarce person is not an AI engineer. Leaders were not asking for developers. They wanted a domain expert with real AI fluency, sitting inside the function — one of three profiles leaders described, and the one most often asked for. One CFO described what he wanted as an accountant by trade who is also an AI-enabled technologist. A CPA doesn’t need to become a developer to be the person in demand.
They can’t find that person, and some have stopped waiting. One CFO at a company of about fifty people spent twelve months searching for a senior hire in tune with the new world and couldn’t find one. A co-founder at a venture-backed fintech went the other way: experience be damned, if a candidate is keen and exploring, that’s a bet worth taking.
What do employers screen for now?
Traits, not credentials. Asked who in their organization was actually succeeding with AI, leaders gave answers that had nothing to do with seniority, age, function or technical background.
Curiosity: Willing to poke at a tool for an evening.
Persistence: Hits a wall and comes back the next day. One leader added this as a second gate: plenty of people are curious, far fewer will do the reps when it gets tedious.
Comfort with ambiguity: Able to work without a defined path.
Functional depth: Knowing the process well enough to see where AI helps and where it doesn’t.
The last one matters more than most candidates expect. Functional depth reads the same on paper whether someone lived the work or watched it, which is precisely why interviews are getting more behavioural in this area.
What questions are interviewers asking?
Not whether you use AI. How.
Three that are becoming standard, and worth preparing real answers for:
- What have you taught yourself in the last six months that nobody asked you to? This is the curiosity test, and it doesn’t have to be about AI.
- What have you built that broke, and what did you do next? This is the persistence test. The failure is the point.
- How do you use AI in your own work? The tell isn’t a good answer, it’s a specific one.
A candidate who describes a messy thing they tried on a weekend is a stronger signal than one with a rehearsed answer about prompt engineering. Interviewers are listening for detail that couldn’t be invented.
Is AI replacing finance and accounting jobs?
The 2026 evidence says no. It says the market is repricing, not shrinking.
The jobs panic cooled during the year. Sam Altman said in May 2026 he was delighted to have been wrong about entry-level white-collar losses. Dario Amodei reframed automation as a multiplier rather than an eraser. The Yale Budget Lab found no meaningful change in unemployment for AI-exposed workers through early 2026.
But the premium moved. A VP of advanced analytics at a large medtech company put numbers on it: in his read, the value of someone who adapts moves from roughly $150K to $250K. He gave one builder a $35K raise to stop a frontier lab from taking them. About nine people left his organization for not adapting, not for cost.
That is the only quantified workforce impact across the sixteen interviews, and it points one way. Headcount isn’t vanishing. The gap between adapters and non-adapters is widening, and it is measurable in dollars.
Canadian pay data for these roles doesn’t exist yet in reliable form. US benchmarks for the technical end are public and steep. For most finance and functional professionals in Canada, that means negotiating without a reference point, on roles that are demonstrably repricing.
How do you prove AI skills without a certificate?
Build a record of work you’ve actually done, written so someone else can judge it. Four prompts per entry:
- What was broken: The specific problem, in the language of your function.
- What you built: The workflow, the tool, the prompt chain, the process change.
- What changed: Time, accuracy, capacity, or work that now happens that didn’t before.
- What you’d do differently: The part that proves you understand the limits.
One worked example beats five vague ones. A month-end reconciliation that dropped from three days to four hours, described precisely enough that a controller can picture it, does more than any list of tools. If you don’t have an example yet, start with work that has the shape the research kept finding: messy data in, judgment out, and a backlog nobody has the hours for.
Keep it to a page. Bring it to interviews. It answers the question leaders said they’re now asking, which is not “do you use AI” but “show me.”
What not to do
Don’t list tools: Naming ChatGPT, Copilot and Claude tells a hiring manager nothing about whether you can direct them.
Don’t claim fluency you can’t defend in conversation: Interviewers are probing this now, and the follow-up is always specific.
Don’t wait for permission: The people leaders singled out built things on their own time before anyone asked.
Don’t assume it has to be technical: The most in-demand profile is the person who knows the work, not the person who knows the model. Depth in accounting, finance, HR, IT or operations is the harder half of the combination to acquire.
Don’t count yourself out if you haven’t started: Curiosity plus functional depth is the profile. The AI part is learnable. A track record of teaching yourself things is a stronger signal than fluency with no functional grounding.
Read the full research
What We Heard, and What the Frontier Is Saying is Clarity’s full report on where AI is working inside companies, where it’s stuck, and the roles leaders are trying to fill. It’s the same research these findings come from, including what leaders said about hiring, screening and what they’re willing to pay for.
About this research
Clarity Recruitment is a specialist recruitment firm in Toronto working with mid-market and growth-stage companies across accounting, finance, HR, IT and operations. The listening tour was conducted between April and June 2026 across sixteen organizations. All interviews were off the record and every finding published here is aggregated and anonymized.
Frequently asked questions
How do I show AI skills on a finance résumé?
Describe work, not tools. For each entry, name what was broken, what you built, what changed, and what you’d do differently. Specific outcomes are the signal hiring managers are looking for now that AI writes a growing share of résumés.
Do accountants need to learn to code to work with AI?
No. The profile leaders described most often was a domain expert with AI fluency, not a developer. Deep functional knowledge plus the ability to design an AI-enabled workflow inside that function is the scarce combination.
What AI questions do interviewers ask in finance roles?
Increasingly they ask how you use AI rather than whether you’re familiar with it, what you’ve taught yourself recently that nobody asked for, and what you’ve built that broke. Expect specific follow-up questions on any workflow you claim.
Is AI reducing finance and accounting jobs?
The 2026 data does not show that. The Yale Budget Lab found no meaningful change in unemployment for AI-exposed workers through early 2026. What the research shows is repricing: people who adapt become more valuable, and people who don’t become more exposed.
What is a Functional AI Expert?
A domain expert in accounting, finance, HR, IT or operations who can design AI-enabled workflows inside their own function. Across sixteen senior-leader interviews, this was the single most requested profile, named unprompted in more than ten conversations.
What do employers pay for AI-enabled finance talent in Canada?
Reliable Canadian benchmarks for these roles don’t exist yet. US figures for the most technical profiles are public and high. The practical implication is that Canadian candidates are negotiating without a reference point on roles that are demonstrably repricing.
