Recruitment has changed more over the past two years than in the previous ten. AI has stopped being an experiment that HR teams tried "next quarter." In 2026, it's part of everyday work — from the first contact with a candidate all the way to signing the contract.

This article isn't about hype. It's about what actually works, where AI helps, and where you still need a human.

Automatic CV screening is no longer a topic

Back in 2023, companies bragged that AI could sort 200 resumes an hour. Today, that's a given. The interesting question is different: how exactly does AI sort them, and does it understand context?

Modern tools like Surikado don't just track keywords. They analyze:

• real experience from job descriptions
• the progression of career growth
• consistency between the CV and answers given in conversation
• soft skills hinted at by how a candidate writes

The result is a ranking of candidates that's not just faster, but often better than what a tired recruiter would produce at five in the afternoon.

Conversation instead of a form

The biggest change in 2026 isn't how AI evaluates a CV. It's that the CV is often no longer the first step.

The candidate starts a conversation on WhatsApp or through a voice interface. AI asks about experience, languages, availability, and expected salary. In a few minutes, the company has the same information it used to get only after reading a CV and making a phone call. Except no one had to make that call.

For candidates, this is a huge difference. They don't need a resume. They don't need to sit at a computer. They answer whenever they have time — during a lunch break, on the way home, in the evening.

If you're curious how this works in practice, check out Surikado for Companies.

Voice interviews as screening

The second thing that stopped being an experiment in 2026 is AI voice interviews. It's not about replacing the human recruiter — it's about screening before them.

AI conducts a 5-to-10-minute interview in the candidate's language. It verifies the basics: competencies, motivation, availability, realistic expectations. The recruiter then reads the transcript and decides who to invite for an in-person meeting.

The effect: the recruiter no longer spends 15 hours a week on initial calls where, within the first two minutes, they already know it's not a fit. They spend that time with candidates who have passed basic selection.

Where AI still doesn't help

To be fair — AI doesn't solve everything:

• Deciding on culture fit. AI can suggest. A human should decide.
• Negotiation. Offer, counteroffer, agreeing on terms — that's still a human domain.
• Onboarding. A new employee's first week makes a huge difference, and that can't be delegated to a chatbot.

Practical recommendation

If you're currently handling recruitment by sitting in Excel and sorting emails, start with one thing. Pick the most common position in your company and try filling it through an AI-first process. Don't change anything else.

In a month, you'll know what works.

In three months, you'll know why you never want to go back.