AI in recruitment is like sex in high school: everyone’s talking about it, few are doing it, and even fewer are doing it well. We talk to AI expert Martyn Redstone to get his take. Keep reading for four practical steps for TA leaders to stop talking about AI and start actually doing it. Safely, fairly, and impactfully.
For all the noise about sweeping AI transformation, most recruitment teams are still firefighting, swamped by applications, and paralysed by indecision.
We spoke to a man who’s had both hands plunged into the AI world for years about turning that around. This isn’t another hype piece. It’s a reality check for TA leaders: what’s really blocking AI progress, where the biggest wins are hiding, what best practices look like, and how to build a fairer, faster, better hiring function without losing your human touch.
Watch Episode 15 of The View now or keep reading for four takeaways.
Your problem is capacity, not capability
Most recruitment teams aren’t “bad at recruiting”; they’re drowning.
It takes roughly 8-9 minutes per CV to review, screen, and update your ATS, Martyn’s data shows. Even if you do nothing else – no stakeholder wrangling; no posting jobs; no recruitment marketing; and certainly no strategic projects – that caps out at around 53 applicants a day; 265 a week.
Set that against typical requisition workload and soaring application volumes, the maths simply doesn’t add up. Recruiting this way is – literally – an impossible job. What looks like slow process and poor candidate experience is a throughput constraint, not a skills gap.
This insight reframes how recruitment teams should think about AI. It’s not about grand, all-encompassing transformation. It’s about relieving pressure.
Too many organisations overcomplicate things, stalling in analysis-paralysis while recruiters remain overwhelmed. Think of AI as a capacity unlock, not a silver bullet.
Start where the work is most repeatable, high-volume, and low-risk. Binary screening (location, eligibility, availability, shift fit, salary range) or FAQ chatbots are ideal places to begin.
These aren’t massive process overhauls but they can deflect ~75% of repetitive admin, cutting candidate wait times and freeing recruiters to focus on meaningful, high-touch work.
Design for modern candidates: immediacy, convenience, agency
TA teams often see a shiny new piece of recruitment tech and go hunting for a problem to match the solution. That’s the wrong way around. As Martyn says, much of his consulting work involves talking teams out of AI – because until you’ve defined the problem properly, no technology can solve it.
For most teams, the real challenge isn’t technology. It’s process. Recruitment still overestimates how much real-time human interaction candidates want and underestimates how much control they expect.
People book holidays, buy cars, and run their finances online without talking to anyone. Candidate expectations have evolved accordingly. They want:
- Immediacy – instant acknowledgement and next steps
- Convenience – flexible, asynchronous options outside 9–5
- Agency – choose-your-own adventure style processes
AI can absolutely enable this but only if the foundations are solid. Layering AI over messy processes won’t fix them. Make sure your fundamentals are right:
- Candidate self-service dashboard in your ATS
- Real-time interview scheduling
- Automated, personalised candidate comms (zero ghosting)
- Transparent timelines and clear status updates
- Channel choice: email, SMS, portal, phone
- Asynchronous assessment (video screening; chatbots; tests)
The aim isn’t AI-driven recruitment for its own sake. It’s to deliver a consumer-grade candidate experience: fast, transparent, and on the candidate’s terms.
Governance isn’t optional, and the buck stops with you
The fastest way to destroy trust – with candidates, managers, and regulators – is to let black box AI make or influence decisions you can’t explain. And trust is only the thin edge of the wedge.
Too many teams are already experimenting unsafely, Martyn warns. Uploading CVs into ChatGPT for screening is unethical, at best, and shows blatant disregard for GDPR. That’s not innovation – it’s risk.
Recruiters mustn’t abdicate responsibility for AI to IT. If AI goes wrong, it won’t be treated as a technical failure. It’ll be a hiring failure – and the buck will stop with you.
Every TA leader should be setting governance guardrails now, to ensure AI adoption is responsible, ethical, and safe:
- Data policy: what can and can’t go into external AI tools
- Explainability: transparent rationale for algorithms influencing hiring
- Bias checks: monitoring for bias sneaking into AI recruitment processes
- Human involvement: ensuring automation assists but humans decide
- Transparency: clear comms to candidates about how you use AI
Governance doesn’t need to be heavy bureaucracy. But as responsible users (and people who don’t want to get fined or sued 💅) we need to mitigate AI’s dark side, so this transformative tech genuinely makes hiring better for everyone.
This principle of fair, sustainable AI is the foundation of Tribepad Sidekick, our fully-integrated, ethical AI recruitment partner. Sidekick can save hiring teams ~14 days and~ £10,093 per 100 hires – without sacrificing transparency and fairness.
Stop talking. Start starting.
The AI conversation is packed with impressive ideas that aren’t yet realities for most teams. Agentic AI. Autonomous AI coworkers. Etc.
But the talent acquisition teams winning won’t be the ones making noise about radical transformation. They’ll be the ones who quietly got started, testing where AI can drive measurable, incremental improvements across their biggest pain points.
True AI “agents” that act fully independently are still a way off. What we have now is agentic behaviour – smart automations that follow logic, learn patterns, and take away grunt work. Demystify it, operationalise it, and get moving.
Martyn’s advice:
- Choose a use case: pilot against a small, low-risk well-defined problem area, like deflecting FAQ emails to a chatbot.
- Stay small: one team, one hiring manager, one metric that matters. Don’t overwhelm yourself.
- Track it: define an owner, timeline, governance, and outcome. Share results transparently, wins and lessons.
- Scale gradually: use the data to gain buy-in and budget to add AI into recruitment in other ways. Rinse, repeat.
This approach turns AI from abstract idea into something tangible and trusted – and sidesteps the analysis-paralysis that kills momentum. You’re not reinventing recruitment overnight – just easing one pain point. Then another, and another, until there are none left. 😎 That’s how you make AI stick.
AI won’t replace recruiters – recruiters who use AI will.
The conversation around AI in recruitment can feel overwhelming – part hype cycle, part existential crisis. But Martyn’s take is refreshingly practical. As you’d expect from someone who’s been in the deep end with AI for years.
AI isn’t here to replace recruiters; it’s here to remove friction. To clear the noise, reclaim time, and create fairer, faster, better hiring for everyone.
Start small. Stay curious. Build trust through transparency and governance. And remember: transformation doesn’t happen through theory or fear – it happens through doing.
Tribepad is the trusted tech ally to smart(er) recruiters everywhere. Combining ATS, CRM, Video Interviewing, and Onboarding, our talent acquisition software is a springboard for fairer, faster, better recruitment for everyone.
Trusted by organisations like Tesco, NHS Professionals, and Subway, 30-million people in 16 languages use Tribepad.
