Do you need an AI interview if your ATS already screens with AI?

Five platforms, one acquisition wave

As of 2026, native AI screening is not an edge case in applicant tracking, it is the default. Workday completed its acquisition of HiredScore for a deal reported at approximately $530 million (Workday itself did not disclose the figure), announced 26 February 2024 and closed around a month later, folding HiredScore's AI-powered sourcing and matching directly into Workday Recruiting. Workday then completed its acquisition of Paradox, the conversational AI hiring assistant, on 1 October 2025. SAP completed its acquisition of SmartRecruiters, which markets itself as 'the Recruiting AI Company', on 11 September 2025, integrating it into SAP SuccessFactors.

The pattern is not limited to the giants. Bullhorn, the staffing-industry ATS, ships its own AI matching engine that predicts which candidates are likely to be hired based on historical placement data, drawing on a claimed 22 billion data points across 800 million candidate records. JobAdder, the Sydney-headquartered ATS majority-owned by SEEK since 2020 (after SEEK built its stake from an initial position taken in 2014), built its own 'Adder Intelligence' feature that summarises and rates every candidate against the job description automatically.

Greenhouse is the outlier, and worth naming as one. Its own AI principles state plainly that 'AI and automation can inform, summarize and surface insight, but it is never the final decision-maker,' with human decision ownership as an explicit design requirement rather than a talking point.

What a ranking actually tells you, and what it cannot

An ATS ranking is built entirely from what a candidate chose to write down: a resume, a cover letter, answers to structured knockout questions. That is a real and useful signal, and at the volumes Greenhouse itself has measured, worth taking seriously: applications per job on its platform rose from about 116 in 2022 to about 244 in 2025, across more than 640 million applications at over 6,000 companies. Software that can sort that pile in seconds is solving a real problem.

What it cannot do is create evidence that was never written down. A candidate who has genuinely done the work they claim, who can explain a decision they made under pressure, or who simply writes a weak CV despite being excellent at the job itself, is invisible to a system that only ever reads the document. If your problem is too many documents, an ATS ranking solves it. If your problem is that the survivors all look the same on paper, ranking them again does not help.

The honest case for stopping at the ATS feature

For a large share of teams, the native ATS screening feature is enough, and buying anything more is the wrong call. If your applicant volume is modest, if the ATS ranking already produces a shortlist your recruiters trust, or if the cost and complexity of another vendor outweighs the value of a deeper first round, stopping at the ATS is not a compromise. It is the correct decision for that team.

'One fewer vendor' is a genuine and frequently underrated reason to stop. Every additional tool is a security review, a contract, an integration somebody has to maintain, and a login the recruiting team has to remember.

Where the gap starts costing you something

The moment this changes is when applicants clear the ATS ranking and still look interchangeable, which happens fastest in roles where the real qualifying signal is not written down: care and frontline work, sales, anything where communication or judgement under pressure is the job rather than a credential you can list. At that point, ranking the same pile of resumes again with a slightly better model does not produce new information.

This is also where the Greenhouse nuance matters in practice. Because Greenhouse deliberately keeps its AI advisory rather than decisive, its customers arrive at this point with less automated sorting already done for them, so the case for an interview tends to arrive earlier for a Greenhouse-native team.