I spent last Tuesday afternoon watching a marketing director struggle to assess whether a candidate had the AI skills they claimed on their CV. The interview was going nowhere - a painfully circular conversation about "AI experience" without either party actually defining what that meant.
The director later admitted to me she had no idea how to test for genuine AI marketing capability. "I don't even know what good looks like," she confessed. Thing is, she's not alone.
In 2026, we're facing a bizarre paradox in marketing recruitment. Everyone wants "AI-skilled marketers" but almost no one can articulate what that actually means or how to verify it. From my work deploying LLM systems across different industries, including marketing tech stacks, I've seen firsthand how this disconnect is creating absolute chaos in hiring processes.
The Illusion of AI Competency in CVs
Let's get something straight: most marketers claiming "AI expertise" on their CVs have basically used ChatGPT to write emails and maybe played with a couple of image generators. That's not AI expertise - that's being a functional human in 2026.
The problem is twofold. First, candidates are wildly overstating their capabilities (who can blame them when every job spec demands "AI skills"?). Second, hiring managers have no systematic way to separate the genuine practitioners from the prompt-pasters.
When I ask marketing leaders what they're actually looking for, I get vague hand-waving about "someone who understands AI." But understanding what, exactly? The technical underpinnings of diffusion models? How to craft effective prompts? How to build automated marketing workflows? How to evaluate AI-generated content?
This ambiguity is killing effective recruitment.
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
How to Master 'Tell Me About Yourself' Interview Question: UK Expert Insights
What Employers Should Actually Test For
Based on my experience building ML systems that marketers actually use (not just talk about using), here are the concrete skills UK employers should be assessing in 2026:
1. Prompt Engineering Literacy
Every marketer now needs to understand how to effectively communicate with AI systems. But I don't mean basic ChatGPT usage. I'm talking about structured prompt development that consistently produces usable marketing outputs.
In practical terms, test for:
- The ability to break complex marketing briefs into structured prompts
- Understanding of context window limitations and how to work around them
- Knowledge of when to use chain-of-thought vs single-prompt approaches
- Ability to evaluate and iteratively improve prompt results
The best candidates will understand prompt patterns as transferable skills that work across platforms, not just hacks for specific tools.
2. AI Output Evaluation & Refinement
The most dangerous marketers are those who can't tell when AI is generating plausible-sounding nonsense. By far the most valuable skill is the ability to critically evaluate AI outputs.
A proper assessment should test whether candidates can:
- Spot factual errors in AI-generated content (they're still common in 2026!)
- Identify brand voice inconsistencies in generated copy
- Recognise when an AI system is exceeding its capabilities
- Understand the editing process for different types of content
3. AI Marketing Workflow Design
I've reviewed countless portfolios where candidates claim they "built automated AI marketing workflows" but can't explain their architecture decisions or the limitations they encountered.
To separate real practitioners from the pretenders, ask them to:
- Sketch how they would design an end-to-end workflow for a specific marketing task
- Explain which parts should be AI-driven vs human-executed (and why)
- Identify potential failure points and how they'd mitigate them
- Describe how they measure the success of AI implementations
4. Ethical & Regulatory Awareness
After the disastrous AI marketing campaigns of 2024-25 that violated the UK's Digital Markets Act and the EU AI Act's transparency provisions, no marketing team can afford to hire someone without solid regulatory understanding.
Ask scenario-based questions about:
- Content attribution requirements under current UK regulations
- Disclosure obligations when using AI-generated imagery
- Data usage limitations for training marketing models
- How to ensure AI systems don't discriminate in audience targeting
Bonus points if they can speak to the ICO's recent marketing AI guidance without needing to Google it first.
How Not to Test for AI Marketing Skills
I've sat through dozens of painful interviews where the assessment approach was fundamentally flawed. Avoid these common mistakes:
- Asking candidates to list AI tools they've used (anyone can name-drop)
- Giving vague briefs like "how would you use AI in our marketing?"
- Treating AI as a separate skill rather than an integrated capability
- Focusing on theoretical knowledge rather than practical application
- Asking yes/no questions about complex implementation decisions
Worst of all is when hiring managers try to assess technical AI knowledge they don't possess themselves. I watched one marketing director ask increasingly technical questions until both she and the candidate were completely lost. The role didn't even require technical ML knowledge - she just felt she should ask "tough AI questions".
Building a Practical Assessment Framework
So what actually works? From my experience helping scale-ups build marketing teams with genuine AI capability, here's a practical approach:
-
The Case Study: Present a realistic marketing challenge with messy constraints and ask candidates to outline how they'd approach it with and without AI assistance
-
The Critique: Show examples of AI-generated marketing materials with deliberate flaws and ask candidates to evaluate them
-
The Workflow Challenge: Ask candidates to design a simple marketing workflow that integrates AI and human touchpoints for a specific outcome
-
The Ethical Scenario: Present a borderline case of AI usage and discuss the considerations they'd weigh
What's absolutely critical is that these assessments mirror real-world complexity. If your tests involve neat, contained problems with obvious right answers, you're testing for theoretical knowledge, not practical capability.
The Integration Mindset
The marketers creating the most value in 2026 aren't AI specialists - they're marketing specialists who understand how to integrate AI into comprehensive strategies.
I recently watched a brilliant candidate fail a series of technical AI questions but then demonstrate exactly how she'd used basic AI tools to improve campaign performance by 32% with half the previous budget. The hiring manager nearly missed this gem because he was too focused on technical AI knowledge.
The most valuable candidates can clearly articulate:
- Where AI excels in their marketing process and where it falls short
- How they measure the ROI of AI implementations specifically
- When to use off-the-shelf tools versus custom solutions
- How they upskill themselves as AI capabilities evolve
Structural Changes to Your Interview Process
To accurately assess AI marketing capabilities, most companies need to restructure their interview process:
- Define exactly which AI skills matter for your specific marketing needs
- Design practical assessments that test application, not just knowledge
- Include both AI enthusiasts and skeptics on your interview panel
- Create scoring rubrics that value demonstrated results over tool familiarity
- Test for learning adaptability - the only truly future-proof skill
Be willing to consider candidates who don't tick every AI box but demonstrate superior critical thinking and marketing fundamentals. I've seen companies pass on exceptional marketers because they couldn't name-drop the latest AI platforms, only to hire technically-fluent candidates with weak marketing instincts.
Sometimes the best AI marketer isn't the one with the most AI skills - it's the one who knows exactly when not to use AI at all.
Looking Ahead
By Q4 2026, I expect we'll see marketing roles split more clearly between AI integration specialists and creative strategists who leverage AI. The current approach of demanding universal "AI skills" without specificity is producing poor hiring outcomes across the board.
For recruiters and hiring managers, the opportunity is clear: develop assessment frameworks that actually reveal practical capability, not theoretical knowledge or tool familiarity. Companies that crack this code will build significantly more effective marketing teams than their competitors still stuck in the "must have AI experience" mindset without knowing what they're actually looking for.
And if you're working with a recruitment partner? Make sure they actually understand the nuances of AI marketing skills assessment. Nothing is more frustrating than a recruiter who can't tell the difference between a prompt-engineer and someone who just uses Canva's AI features.
What marketing leaders need now isn't AI specialists - it's marketers who understand how AI changes their discipline, and can articulate exactly where it adds value and where it doesn't.
Some things don't change though - the best candidates are still those who can tell a compelling story, regardless of which tools they use to craft it.
