Is your recruitment firm holding you back on AI?

A client asks to speak on Slack. Getting Slack approved takes a week.

You want a tool to transcribe meetings so you can concentrate on the conversation. That needs approval too. You find something that could save hours of administration, but connecting it to your email, calendar or CRM requires another review.

These are ordinary things recruiters want to do to serve clients and get their work done.

The response I keep hearing from recruiters at the big shops is: “We’re allowed to use Copilot.”

Then they tell me it is next to useless. In the setup their company has given them, it isn’t connected to much of what they actually need. They still have to find the information, copy it across and update the systems themselves.

They have access to AI. They just can’t use it to do much of the work.

Governance on steroids. Paralysed by process.

Meanwhile, smaller private firms are connecting their systems, testing workflows and using them in real searches. If you are a good recruiter stuck waiting for approval, you should be asking what this means for your own career.

What recruiters are actually allowed to do

I see a chasm opening between parts of the public and private recruitment markets.

That is what I am hearing in conversations. It does not mean every public company is slow or every private company is ahead. Large firms have substantial technology budgets, capable people and responsibilities that smaller businesses may never have to carry.

But an investment in AI at headquarters does not tell you much about what a recruiter is allowed to do on Tuesday morning.

Recruiters I speak to in listed companies often describe extraordinary barriers. People in private businesses have more room to experiment, see what works and put it into use.

Having an approved AI subscription is a start. AI becomes much more useful when it can work directly with the information and systems you use every day.

Can it research a client against the actual search brief? Can it use your notes and call transcripts? Can it prepare candidate briefing content and update the pipeline? Can it use what you learned in the client meeting when researching candidates?

Or are you still copying information between windows, asking it to rewrite an email and then returning to the same manual process?

A company can talk enthusiastically about AI while its recruiters remain stuck doing the work much as they did before.

A recruiter working on an RPO assignment recently told me they effectively had access to LinkedIn and a spreadsheet. There was probably some hyperbole in that description, but enough truth to make me stop and think.

This person works for one of the bigger RPO businesses. From the outside, it appears to be ahead of its competitors in the technology available to it. Yet the recruiter’s description of their day-to-day work sounded remarkably familiar to me from when I was running an RPO business.

An RPO recruiter may also be working within the client’s systems and restrictions. That can make change harder. But if the person doing the recruiting still spends their day moving information between LinkedIn and a spreadsheet, how much has actually improved?

The technology a company owns and the technology its recruiters can use are often two very different things.

What we are doing at Aspiring Partners

TalentHub is our own platform. It reads and writes across emails, calendars, notes, call transcripts and files in client and candidate Drive folders as part of our workflows.

We also have our own Mac-native app running a whole raft of processes. It brings together Apple, Google, Anthropic and OpenAI, with Grok used where it makes sense.

Every time a model changes, we test it. We look at what it does well, where it falls short and whether we should change which model runs a particular process.

We do not pick a provider and assume it will always be the best choice. We keep testing against the work we need done.

We have several hundred MCP tools connecting Claude and TalentHub, with more to come. I have lost count of the custom agents we have built.

These workflows add jobs, run client research, create candidate briefing content and add candidates to the pipeline. Information from a conversation can be used in the next stage of a search without someone having to find it, copy it and explain it all over again.

There is still plenty to check, correct and improve. But we are running these processes in our work and learning from the results.

The people using these tools every day are getting better at using them. They are learning what to trust, what to check and where a process needs a person involved.

A recruiter waiting for permission is missing that experience as well as the immediate benefit of the tool.

Managing PCs and a CRM is no longer enough

Many recruitment firms built their IT function around managing PCs, keeping the CRM running, controlling access and maintaining security.

That work still matters. But someone now needs to understand APIs, AI models and agents. They need to know how to connect systems, assess the results and work out why a process has failed.

For some CIOs and IT managers, this is simply outside the skills they have developed.

Others may have the ability but have been told to prioritise security and governance above everything else. If their performance is judged mainly on preventing problems, it is hardly surprising that they become good at saying no.

Who is responsible for helping the recruiter get more done?

Recruitment firms handle information people have entrusted to them. We have to protect it. I am not suggesting we throw away controls or let people put confidential information into whichever tool they happen to like.

But there has to be a practical way to test and improve how we work.

If a small test with clear limits takes months to approve, the company should count the cost of that delay. Other firms are using those months to try things, make mistakes, correct them and get better.

The client who wants to speak with you on Slack does not care how many internal approvals you need.

They care whether working with you is easy and useful.

This affects other industries too. Recruitment is simply where I see it happening.

I don’t want AI sending more recruitment spam

The next step for us is outreach automation.

I have not automated that yet. My focus has been senior leadership roles, where an approach often needs considerable judgment and a personal understanding of the individual.

But as we build teams further down the organisation and take on more individual contributor searches, there is no way in hell I would expect recruiters to run high-volume work manually when we can build a better process.

We are testing how AI could support outreach based on the relationship we have with a candidate. Someone who knows us well should not receive the same approach as someone who has never heard of us. The relationship should affect what we say, how we contact them and whether we should contact them at all.

I believe spam is already damaging the recruitment market. Candidate inboxes are flooded with the same demand for attention:

Job. Job. Job. Job.

Making those messages cheaper to produce will not make them more welcome.

I believe recruitment will increasingly depend on consent and referrals. People choosing to hear from you. Someone they trust introducing you. A recruiter knowing enough about their work and circumstances to make the conversation worth their time.

AI can help us prepare for that conversation, remember what someone has told us and avoid approaching them with something unsuitable.

If we use it mainly to send more poorly judged messages, we will make the market worse.

Good recruiters already understand this. Their relationships are built on knowing when they have something worth discussing, and when to leave someone alone.

What does this mean for your career?

If you are a high-performing recruiter in a large public firm, look seriously at whether your company is helping you keep up.

You may have strong client relationships, deep domain expertise and candidates who take your calls because they trust you. You may also be spending too much of your working life on processes that another recruiter is already learning to automate.

Good billings today do not mean you can afford to ignore that.

Ask your firm for something concrete: an approved environment in which you can test a real workflow, someone who can help connect the systems, and time to learn how to use it properly.

Then look at what happens.

Does the company help you move forward? Does it explain the concerns and help you work through them? Or does every proposal disappear into another committee or approval process?

You do not need to become a software engineer. You do need to understand enough to know what is possible and what support you should expect.

There are good reasons to work for a large recruitment company. It may give you access to clients you could not reach alone, a trusted brand, capable colleagues, training and reliable operational support.

A good firm should be able to explain how it makes you better at your job.

But if you bring the client relationships, the candidate trust and the domain expertise, and the firm keeps preventing you from learning how to work more effectively, then you need to ask yourself:

What is the company actually providing?