What is an AI recruiter? What it does, what it should not do and the rules

An AI recruiter is software that uses AI models and automation to carry out parts of a recruiter's job: finding candidates, writing outreach, answering questions, screening applications against requirements and booking interviews. The term is also used for human recruiters who work heavily with AI tools. In practice, today's AI recruiters handle the repetitive, high-volume steps, while people stay in charge of interviews and hiring decisions.

This guide explains what AI recruiters actually do, where they help and where they cause problems, how to evaluate AI recruiting software, and which rules apply in the EU and the US.

What an AI recruiter does

Most products combine a few of these capabilities. Few do all of them well.

• Sourcing: searching databases and profiles, or running targeted ads, to find people who match a role
• Outreach: drafting and sending personalised messages and follow-ups
• Conversational screening: asking candidates qualifying questions by chat, text or voice, at any time of day
• Application screening: comparing CVs or answers with the requirements and ranking or filtering candidates
• Scheduling: offering interview slots and booking them in the hiring manager's calendar
• Admin: summarising conversations, updating the ATS and sending status updates

Where AI recruiters help most

The value is greatest where volume is high and the steps are repetitive. A warehouse, retailer or care provider hiring dozens of people a month can lose good candidates simply because nobody replies for three days. An AI assistant that answers within minutes, checks the basics such as right to work, licence, shifts and location, and books an interview, keeps those candidates in the process.

• High-volume hourly roles with clear, checkable requirements
• Out-of-hours applications, when candidates expect an instant reply
• Interview scheduling across many managers and sites
• Keeping the ATS up to date without manual data entry

What AI recruiters should not do

AI is weak where judgement, context and fairness matter most. Models can repeat patterns from historical data, including biased ones, and they can reject good candidates whose CV does not look like the typical profile. Keep people responsible for the decisions that affect someone's career.

• Make final hiring or rejection decisions on its own
• Score personality, emotion or honesty from video or voice
• Reject candidates on criteria you cannot explain to them
• Replace the conversation in which a candidate decides whether they want the job

AI recruiter vs human recruiter: who does what

The most useful way to think about an AI recruiter is as a split of tasks, not a replacement. A sensible division for most teams looks like this.

If a vendor's demo shows the AI making the final call, treat it as a warning sign, both for quality and for compliance.

AI recruiting software: the main types

The rules: EU AI Act and US laws

Recruitment is one of the most regulated uses of AI. In the EU AI Act, AI systems used to place targeted job ads, analyse and filter applications or evaluate candidates are classed as high-risk. The Digital Omnibus adopted in 2026 moved the main high-risk obligations for these systems to 2 December 2027. They include risk management, data quality, documentation, human oversight and information to the people affected. The ban on emotion recognition in the workplace already applies.

In the United States, New York City's Local Law 144 requires employers who use automated employment decision tools to have them independently audited for bias every year, to publish a summary and to notify candidates. Other states and cities have their own rules, and anti-discrimination law applies to AI decisions just as it does to human ones.

• Ask the vendor how the system is tested for bias, and for the results
• Make sure a person reviews every rejection the system suggests
• Tell candidates when AI is used and how they can reach a human
• Keep records of how the tool is configured and what it decided

How to evaluate an AI recruiter

• Start with the bottleneck. Is your problem too few candidates, too slow screening or too much scheduling? Buy for that
• Test it on one real role, with real candidates, for a few weeks
• Compare time to first interview, share of candidates who reach interview and candidate feedback before and after
• Check that it writes into your ATS with the source recorded
• Ask where candidate data is stored and how long it is kept
• Read the candidate side yourself. Apply to your own job and see how it feels

Questions to ask AI recruiting vendors

• What exactly does the AI decide, and what does a person decide?
• How was the model trained, and how is it tested for bias across gender, age and ethnicity?
• Can we see why a candidate was ranked or filtered out?
• How do candidates know they are talking to AI, and how do they reach a human?
• Where is candidate data stored, who can access it and when is it deleted?
• How does the product meet the EU AI Act high-risk requirements, and what is the plan for December 2027?
• Which ATS integrations are live today, not on the roadmap?

AI does not create candidates

An AI recruiter can screen and schedule faster, but it can only work with the people who are already in your pipeline or in the databases it can search. For roles with a shortage of active job seekers, the limiting factor is reach, not processing speed. That is where targeted advertising to passive candidates comes in, with AI used to pre-screen the people who respond.

How SmartHire uses AI

SmartHire runs targeted campaigns on Facebook, Instagram, TikTok and Snapchat to reach people with the right experience, including those who are not job hunting. AI is used to screen responses against your requirements, and only candidates who match are delivered into your ATS. Your team runs the interviews and makes the decisions. For most roles the first qualified candidates arrive within two to five days.

What is an AI recruiter?

Software that uses AI and automation to source, screen, message and schedule candidates on behalf of a recruiting team. The term is also used for human recruiters who rely heavily on AI tools.

Can an AI recruiter replace human recruiters?

Not fully. It can take over repetitive steps such as first screening and scheduling, but interviews, judgement and hiring decisions should stay with people, and the law in many places requires human oversight.

Is AI candidate screening legal?

Yes, with conditions. In the EU it is classed as high-risk under the AI Act, with obligations applying from December 2027. In New York City, automated tools need a yearly bias audit and candidate notice. Anti-discrimination law applies everywhere.

What is the difference between AI recruiting software and an ATS?

An ATS stores and manages candidates and the hiring process. AI recruiting software adds automation on top, such as sourcing, screening or scheduling, and often writes its results into the ATS.