Most hiring managers are testing creative director candidates on the wrong thing. They ask which AI tools a candidate uses, how fluent they are in Midjourney or Firefly, whether they've run a generative workflow at scale. Tool fluency is the easiest thing to fake and the quickest to date. What predicts a good creative director hire in 2026 is judgement: whether the candidate can decide what the machine is allowed to touch, where it must stay out, and who answers when it gets it wrong.
That's a leadership question, and it needs a leadership assessment. If your process for creative director hiring still ends with a portfolio walk-through and a chat about culture fit, you're measuring taste but not the thing that's changed.
Why tool fluency tells you so little
The tools turn over every few months. A candidate who's brilliant with this quarter's image generator may be useless with the next one, and a candidate who's never touched it can be up to speed in a fortnight. Nobody pays a creative director for button knowledge.
I spent twelve years in consumer PR watching what happened when a brand's voice got diluted. A campaign would go out slightly off, a tone that was almost right, and the press noticed before the client did. Journalists are very good at smelling copy that nobody cared about. Generative output has made that failure cheaper to produce and harder to spot internally, because the work looks finished. It's polished and plausible and it belongs to nobody.
So the question for a candidate should be "what do you do when the tool gives your team something good enough?" rather than "can you use these tools?" Good enough is the dangerous outcome. Terrible output gets binned. Competent output ships.
Beyond Tick Boxes: Diversity Recruitment Strategies That Actually Transform UK Workplaces
Master the Virtual Hot Seat: 7 Video Interview Techniques Recruiters Don't Tell You
AI Governance Careers Surge in 2026: Why This Role Will Define the Next Decade
What brand authenticity looks like as a hiring test
Ask the candidate to describe a brand they've led and to name three things that brand would never do. Specific things, not values-deck phrases. Then ask which of those three a generative tool would break first, and what they'd put in place to stop it.
Strong candidates answer this quickly because they've already thought about it. They'll talk about reference libraries, house rules for prompts, a named person who signs off anything with synthetic imagery in it. Weak candidates talk about "guardrails" in the abstract and then drift back to speed and efficiency.
A set task works better than an interview for this. Give them a real brief with a thin brand guide and a generated first draft that's subtly off-tone, and ask them to mark it up. You're watching what they notice. Can they say why a headline sounds like every other headline? Do they catch the legal exposure in the claim, or the image that quietly borrows a competitor's look? I've seen senior people wave through work like that because it arrived fast and looked clean.
Testing how they run a team that's using the machine
This is the part most processes skip. A creative director doesn't make most of the work, they set the conditions under which other people make it. So ask about the juniors.
What happens to a junior designer's development when the first draft is generated for them? It's a serious question and a candidate who's never considered it hasn't managed a team through this period. Good answers involve deliberately protecting some work from automation so that people still learn to see. Others involve rotating who critiques the output, or banning generated material at the concept stage and allowing it at production. You don't need to agree with the answer. You need to hear that there is one.
Ask, too, about disclosure. Who tells the client when generative tools were used, and when? With copyright and data protection questions still unsettled, a candidate who treats this as someone else's problem is a risk. The ICO's guidance on AI and data protection is a reasonable thing to expect a serious candidate to have at least skimmed.
For what the portfolio side of the conversation looks like now, our earlier piece on design portfolio trends and what creative directors want to see is worth reading before you brief an interview panel.
The strongest objection
The fair counter-argument is that judgement is a soft thing to assess, and that a hiring manager under time pressure will reach for something measurable. A tool test has a right answer. A conversation about taste doesn't, and different panellists will score it differently. That's a real problem, and I won't pretend structured scoring fully solves it.
But the alternative is to measure the wrong thing precisely, which is worse. And the judgement questions aren't as vague as they sound. A candidate either names the specific failure modes of their last brand or they don't. They either have a sign-off process or they don't. Put those on a scorecard with plain yes/no anchors and you've got something repeatable enough to compare candidates, without pretending to a precision you haven't got.
If you're recruiting at this level and want a shortlist that's already been screened for this kind of thinking, OHUB's Elevate service handles senior creative searches, and the employer side of the platform lets candidates show their reasoning on video rather than in a covering letter, which suits this problem well.
What to change before your next search
Rewrite the brief for the role. If the job description lists tools, strike them out and replace them with accountability: who owns brand voice when output is generated, who approves what goes to a client, who trains the team. Candidates read these documents closely and the good ones will self-select.
Then ask yourself something uncomfortable about your own panel. Would your current interviewers recognise a strong answer on AI governance if they heard one?
