AI is everywhere, from creation of marketing visuals to conception of engineer features, and now, in recruitment…screening interviews hosted by AI! 

Sounds great, right? But there’s one problem we see companies overlook:

AI can make a good hiring process faster. It can also make a bad hiring process faster.

If your hiring criteria are unclear, your job description changes every few days, interviewers are asking completely different questions and hiring managers take a week to provide feedback, adding an AI recruiting tool won’t solve the problem.

Before investing in AI recruitment tools, growing companies should first make sure they have a solid hiring foundation.

The real problem might not be recruitment

When a company struggles to hire, the first instinct is often to look for a better sourcing tool or a more powerful applicant tracking system and why not have AI do the screening in hopes to filter the best candidates?

But sometimes, the problem starts much earlier.

We’ve seen hiring processes where:

• The job description changes after every candidate interview.
• Three interviewers have three different ideas of what a successful candidate looks like.
• Hiring managers add new “must-have” requirements halfway through the search.
• Candidates go through multiple interviews without knowing what each one is assessing.
• Interview feedback sits for days waiting for a decision.

None of these problems are solved by AI. They are process problems.

Start with the hiring criteria

Before you post a job, everyone involved in the hiring process should agree on what you’re actually looking for.

This sounds obvious, but it’s often where things start to go sideways.

A hiring manager might say they need a senior person with strong leadership skills.

But what does “senior” mean? Does the person need five years of experience? Experience managing people? Experience working in a startup? The ability to work independently? Experience with a specific technology?

And what does “strong leadership” actually look like in the role? Instead of relying on vague requirements, define the competencies and outcomes that matter.

For example, instead of “Strong communicator”, define it with “Can communicate complex technical information clearly to non-technical stakeholders and adapt communication to different audiences.”

Now you have something you can actually assess during an interview.

Then, use AI where it saves time

Once your process is structured, there are plenty of recruitment tasks where AI can genuinely help.

Job descriptions

• AI can help turn your role requirements into a clear, engaging job description.
• It can improve structure, suggest more inclusive language and adapt the tone to your employer brand.
• But don’t ask AI to decide what the role should be.

The hiring team defines the role. AI helps communicate it.

Candidate communication

Recruitment involves a surprising amount of repetitive communication.

Interview invitations. Follow-ups. Reminders. Status updates. Rejection emails.

AI can help recruiters draft these messages quickly and maintain a consistent tone. Meanwhile recruiters more time to focus on the conversations that actually require a human.

Interview questions

Once you’ve defined the competencies you’re looking for, AI can help generate behavioural and situational interview questions.

For example, if you’re assessing stakeholder management, AI can generate several questions designed to uncover how a candidate has actually handled difficult stakeholders in the past.

But the hiring team still needs to decide whether those questions are relevant to the role.

Interview summaries

AI can also help organize interview notes and summarize candidate responses.

This can be particularly useful when several people are involved in the hiring process.

Instead of reading through pages of notes, a hiring manager can review a consistent summary organized around the competencies being assessed.

But there’s an important distinction:

AI can organize the evidence. It shouldn’t make the hiring decision for you.

What about AI screening candidates?

Don’t forget about the hiring manager

There’s another part of recruitment that AI can’t fix: manager accountability.

You can have the best recruiting technology in the world, but if your hiring manager takes five days to review resumes, cancels interviews at the last minute or can’t agree on what they’re looking for, your process will still be slow.

That’s why hiring managers need clear responsibilities.

• Before the search, agree on the role, competencies and interview process.
• During the search, provide timely feedback and stick to the agreed criteria.
• After interviews, evaluate candidates based on evidence rather than instinct alone.
• At the decision stage, make a decision within an agreed timeframe.

Recruitment isn’t just an HR responsibility. Hiring managers are part of the recruitment process and they need to own their part of it.

So, can AI improve recruitment?

Absolutely.

But the companies that get the most value from AI in recruitment aren’t necessarily the ones with the most AI tools. They’re the ones with the clearest hiring processes.

They know what they’re hiring for. Their managers are aligned. Their interviews are structured. Their candidates know what to expect. Their decisions are made quickly.

Then they use AI to remove the repetitive work around that process. That’s when technology starts making a real difference.

Fix the process first. Add AI second.

Is your hiring process ready for AI?

Before investing in another AI recruiting tool, ask yourself:

• Is our job description clear and stable?
• Do we know what success looks like in the role?
• Are our hiring criteria based on competencies rather than vague requirements?
• Are interviewers assessing the same things?
• Do we use structured interview questions?
• Do hiring managers know their responsibilities?
• Do we have clear timelines for feedback and decisions?
• Is our candidate experience consistent?
• Do we know which parts of recruitment are actually taking too much time?

If the answer is “not yet” to a few of these questions, you might not need more AI.

You might need a better hiring process first. And that’s a much better place to start.