You’d be hard pressed to find a software business in the world right now not using AI to some degree.
Cars have AI. Golf clubs have AI. Your parents are using it to find recipes. Your kids are using it to “help” them with their homework. It’s already touching every part of the world in a big way.
So how much should you focus on it as an agency?
Should you use an AI chatbot on your website to enhance the user experience for candidates? Or simply replace every single human in your business with OpenAI Premium?
I spoke with someone from another software business last week who told me potential clients were asking them on sales calls whether they could strip back their subscription if they let half their team go at some point.
So recruitment leaders are already thinking about AI to the point of analysing their SaaS expenditure. Which makes me think ‘voluntary’ redundancies aren’t as far away as we think.
But amid the constant pressure to keep up with the Jones’, to what extent should you follow the trend? And how much AI is too much AI? And what if pursuing automation risks eradicating empathy and nuance in relationship building?
The question isn’t whether you should use AI any more. It’s where it should stop.
We’re here to find out.
What can you use AI for?
Yes, AI’s great for creating deepfakes of your Finance Director swinging from the chandelier at the Christmas party, but there are also other uses. I can’t promise they’re as entertaining, but they might make you more money.
Some have defined the clever use of AI as the difference between smart automation and lazy substitution.
You might use it for:
- Admin & Ops
- Content & Marketing
- Candidate sourcing
- Training & development
- Relationship building
- Negotiation
- Candidate interaction
And for each of these, let’s look at the specific activities you might substitute.
For admin & ops you might think about data cleansing, meeting summarisation, calendar management, or updating the CRM or database.
For content & marketing you might look at job ad writing, with tools like Mitch’s AI, email nurturing or even report writing for the industry.
For candidate sourcing you could look into AI CV matching or even analysis during early stages of the process.
For training & development, could you use an AI coach or knowledge assistants for onboarding juniors.
All of the above you could argue have positives and negatives. Does an AI coach have the nuance to develop a senior recruiter’s skills and abilities? Does it take the human out of it? Would a junior recruiter enjoy training with a robot? Or find it too hands-off and synthetic?
You could argue the last three on the list are the least likely places AI might help.
For relationship building, on the face of it, AI is less capable than real people. Your clients and candidates buy from a place of trust. And AI’s not as trustworthy as a human being.
But small touch points over email or text, designed and calibrated by AI, that prompts human recruiters to do what they do best? That might be worth thinking about.
For negotiation? It’s highly unlikely AI or an LLM will replace human recruiters. Even if they’d be more effective, putting your clients through their paces against a bot might be a step too far. You’ll also be hard pressed to justify 25% if you’re not paying the bots you ‘employ’.
And candidate representation? Humans are still number one in phone calls and in-person interaction. But is there an argument for first or technical interviews being conducted by AI? I think so.
I’m also not sure candidates trust AI any less than recruiters when it comes to spotting technical ability.
So, what could you use AI for? It depends. On you, your business, the technology and the strengths and weaknesses of your team.
The business case for AI in recruitment
McKinsey, Deloitte and PwC all estimate 25–30% of HR/recruitment tasks could be automated by 2030.
Weirdly, that seems light for me. And it’s also not to say, following that adoption, 30% of all recruitment staff will be redundant. It may just rid existing recruiters of the admin they currently do. The manual busy-work that fills their valuable time with admin, instead of what they’re good at.
Right now AI can easily reduce admin time in CV formatting, interview scheduling, or updating the database.
It can improve candidate search breadth and accuracy.
You can use it to create better reporting and forecasting.
And when applied thoughtfully, it should multiply your output rather than replacing the people who’ve got you this far.
According to some statistics, those agencies that adopt automation might find 20–40% faster time-to-fill rates and higher employee satisfaction.
The risk of over-automating
Whether you use AI in a client-facing role, or just behind the scenes, there’s a very real possibility of overusing it.
Your CFO will love the idea of getting rid of junior members of staff off the payroll. They cost a lot to train, and for 50% of them you’ll never recoup the outlay.
But those who work out? They become your top billers of tomorrow. Only, if they’re never given a chance to cut their teeth, they may be loyal to those who gave them the chance.
Equally, should you decide to focus AI on outward facing responsibilities, it doesn’t take a wild imagination to see how it could backfire. Some clients might like the efficiency. Some may find the lack of human touch distasteful.
Think about walking into a hotel. How does having a self-check-in screen make you feel over a human receptionist? Where one feels premium, where the other feels low-end.
The key with AI in recruitment might be about evolving slowly.
For example, using it, not for killing the junior recruitment roles you have, but killing the low-value junior recruitment work. And allowing those in junior positions to focus on high-touch work you’d associate with senior consultants and a high-end delivery.
The future of agency recruitment
Any guesses about agency recruitment of the future are likely to be wrong. They’ll also rely on cliche and stereotype.
Future industry leaders are likely to be running companies that are hybrid systems of bot and human.
Humans placed the front, building relationships, giving advice and working with empathy. But with an AI undercarriage helping your speed to market, the accuracy of your work, and the depths of your data.
Those at the top will see their tech as invisible. They’ll probably call it their difference maker. It might have a name that differs to others, but underneath the same engine will be used in different ways.
We’ve seen this in other industries already. In 2025 you find accountants propped up by software which gives them their USP. Or marketers who use automation, so they fire on all cylinders.
It won’t be long until recruitment follows suit, if it hasn’t already.
So how should you manage the human side of AI?
Well the fact you’re already asking this question shows you’re aware of the disruption that could happen. And your recruitment business is more human-centric than most.
Your products are human. As is your customer base and staff.
If you’re about to ask ‘can I drop down my SaaS subscription if I let go half my team?’ it’s likely you’re not thinking about the long-term impact of using AI in your business. You might be at risk of going too far, too soon.
The question obviously isn’t “should we embrace AI?”
You should. And if you don’t you’ll be left behind. But swapping it for the humans in your company is risky business. You might find it’s a mis-step that’s difficult to walk back from.
Your goal should be augmentation over replacement.
Recruitment is still, and will always be, about people. The closer AI can bring you to those people, the better off you’ll be.
