Key takeaways
- Greenhouse leads the AI applicant tracking systems at 67/100, ahead of PageUp (64) and SmartRecruiters (63). They are the only three of 52 to reach a D; no ATS reaches a C.
- What gets regulated is the AI inside the ATS - ranking, matching, screening and now AI interviewers - not the ATS itself. Score each AI feature you plan to switch on, not the platform.
- Bias audits are the widest gap: 27 of the 52 score 30 or below on bias audit transparency, and the leaders' audits mostly run on test data. Ashby's audit of its Criteria Evaluation on 3.6 million real applications is an exception.
- No ATS scores above 45 on FRIA support, so EU employers will be writing their own fundamental rights impact assessments.
- Several vendors' compliance pages still cite Colorado's repealed 2024 Act or the EU AI Act's old August 2026 date. Check which law a vendor's statement was written for.
An AI applicant tracking system is an ATS with AI built in: matching and ranking applicants against a job, screening resumes, writing summaries, and - new in 2026 - running first-round interviews with an AI agent. Every major ATS now ships some of this. What separates them is not the feature list but the evidence a buyer can check: a bias audit with real numbers, documentation of how a score is built, and controls that keep a person in the decision.
We score every vendor on the same rubric, from public evidence only. Below are all 52 applicant tracking systems in the directory, ranked by score.
How we ranked them
Each ATS gets a 0-100 score in seven categories, weighted into one overall score:
| Category | Weight | What it asks |
|---|---|---|
| Article 11 technical documentation | 20% | How the AI features work, documented |
| Bias audit transparency | 18% | Published audits, auditor, data, cadence |
| FRIA support | 15% | Material a deployer needs for an impact assessment |
| Data governance disclosure | 15% | Training data, retention, subprocessors |
| Human oversight design | 12% | Overrides, switches, audit trails |
| Post-market monitoring | 12% | Change logs, incident handling, ongoing testing |
| Customer documentation | 8% | Deployer guidance, notices, instructions for use |
One point shapes every score here. The ATS itself is a system of record; what the EU AI Act, NYC Local Law 144 and the state laws regulate is the AI surface inside it - the features that score, rank or screen candidates. So the evidence that counts is evidence about those features. A score is a statement about that evidence, not about whether the ATS is good software, and none of these is an endorsement.
This ranking covers tools filed under ATS as their primary category. The ATS category page also lists tools that offer ATS features as a secondary line, such as Eightfold AI (67) and Alva Labs (61).
AI applicant tracking systems at a glance
| Rank | ATS | Score | Bias audit (of 100) | Last reviewed |
|---|---|---|---|---|
| 1 | Greenhouse | 67 (D) | 84 | 2026-11-14 |
| 2 | PageUp | 64 (D) | 80 | 2026-09-25 |
| 3 | SmartRecruiters | 63 (D) | 80 | 2026-11-14 |
| 4 | Ashby | 59 (F) | 72 | 2026-11-14 |
| 5 | Workday Recruiting AI | 59 (F) | 55 | 2026-11-14 |
| 6 | Dayforce | 57 (F) | 42 | 2026-08-02 |
| 7 | Pinpoint | 57 (F) | 42 | 2026-07-17 |
| 8 | Cornerstone OnDemand | 57 (F) | 45 | 2026-11-14 |
| 9 | Hireology | 57 (F) | 83 | 2026-09-25 |
| 10 | Rippling | 56 (F) | 42 | 2026-07-17 |
| 11 | Tribepad | 56 (F) | 78 | 2026-09-25 |
| 12 | Lever | 56 (F) | 62 | 2026-11-14 |
| 13 | SAP SuccessFactors Recruiting | 55 (F) | 40 | 2026-11-14 |
| 14 | iCIMS | 53 (F) | 50 | 2026-11-14 |
| 15 | Jobvite | 53 (F) | 52 | 2026-11-14 |
| 16 | Bullhorn | 52 (F) | 42 | 2026-11-14 |
| 17 | Workable | 51 (F) | 38 | 2026-11-14 |
| 18 | Oracle Recruiting | 51 (F) | 28 | 2026-11-14 |
| 19 | Paycom | 50 (F) | 38 | 2026-07-17 |
| 20 | UKG | 50 (F) | 28 | 2026-11-14 |
| 21 | Avature | 48 (F) | 40 | 2026-09-25 |
| 22 | ADP | 48 (F) | 46 | 2026-07-17 |
| 23 | Oleeo | 48 (F) | 40 | 2026-11-14 |
| 24 | Jobylon | 48 (F) | 28 | 2026-09-25 |
| 25 | isolved | 47 (F) | 45 | 2026-09-25 |
| 26 | Teamtailor | 47 (F) | 30 | 2026-07-17 |
| 27 | HiBob | 46 (F) | 20 | 2026-09-25 |
| 28 | Factorial | 46 (F) | 25 | 2026-09-25 |
| 29 | ClearCo (ClearCompany) | 46 (F) | 28 | 2026-09-25 |
| 30 | Darwinbox | 46 (F) | 28 | 2026-09-25 |
| 31 | JobAdder | 45 (F) | 35 | 2026-09-25 |
| 32 | Avionté | 45 (F) | 22 | 2026-09-25 |
| 33 | Cegid | 45 (F) | 30 | 2026-09-25 |
| 34 | Vincere | 44 (F) | 18 | 2026-09-25 |
| 35 | JazzHR | 44 (F) | 35 | 2026-08-02 |
| 36 | BambooHR | 43 (F) | 28 | 2026-07-17 |
| 37 | Breezy HR | 43 (F) | 25 | 2026-08-02 |
| 38 | Paylocity | 43 (F) | 33 | 2026-07-17 |
| 39 | Recruitee | 42 (F) | 30 | 2026-07-17 |
| 40 | softgarden | 42 (F) | 22 | 2026-09-25 |
| 41 | Zoho Recruit | 42 (F) | 24 | 2026-07-17 |
| 42 | Cezanne Recruitment | 41 (F) | 24 | 2026-09-25 |
| 43 | Harri | 41 (F) | 22 | 2026-08-02 |
| 44 | Recruitly | 41 (F) | 15 | 2026-09-25 |
| 45 | Manatal | 41 (F) | 22 | 2026-07-17 |
| 46 | Recruit CRM | 40 (F) | 20 | 2026-09-25 |
| 47 | Personio | 40 (F) | 25 | 2026-08-02 |
| 48 | Paycor | 40 (F) | 25 | 2026-09-25 |
| 49 | Keka | 37 (F) | 20 | 2026-09-25 |
| 50 | Recruiterflow | 36 (F) | 18 | 2026-09-25 |
| 51 | Ceipal | 35 (F) | 20 | 2026-07-17 |
| 52 | JobDiva | 29 (F) | 15 | 2026-09-25 |
Tools that share a rounded score are listed in order of their unrounded scores: Ashby is just ahead of Workday Recruiting AI at 59, and Dayforce, Pinpoint, Cornerstone and Hireology all round to 57, in that order.
1. Greenhouse - 67/100
Greenhouse is one of the most established applicant tracking and structured-hiring platforms. Its AI includes Talent Matching, which sorts applicants into Strong, Good, Partial or Limited match against a recruiter's weighted criteria, resume anonymisation, scorecard summaries, a notetaker, and Voice AI, the AI interviewer it gained by acquiring Ezra AI Labs in May 2026.
Strengths. The best evidence of any ATS on technical documentation and bias audits. It publishes its ISO/IEC 42001 certificate, monthly Warden AI audits of Talent Matching and the AI interviewer, a public AI features index naming 17 features with their model providers, and a status page that tracks Voice AI. No mechanism advances or rejects a candidate on an AI rating, and any Voice AI score override is logged with a written rationale.
Watch out for. The audits run on Warden's test data. Voice AI's optional communication score - clarity, fluency, confidence and engagement, worth up to 20% of the overall score - has no audit, and the product page says the interviewer never scores fluency or confidence. Voice AI sits outside the current ISO 42001 certificate, and there is no FRIA template. Thinking of switching? See Greenhouse alternatives.
Best for: teams that want the best-documented AI layer in the category and will keep Voice AI's communication score switched off.
2. PageUp - 64/100
PageUp is an enterprise talent acquisition suite, based in Melbourne. Its main AI feature, AI Skills Matching, uses an LLM to pull skills from the job and the resume, then a fixed rule-based algorithm sorts each applicant into one of five bands. Paige, its conversational assistant, went on general sale in June 2026.
Strengths. A monthly Warden AI LL 144 report, the disclosures Colorado's SB 26-189 asks of developers already published, and clear guidance telling NYC customers not to make Skills Matching the sole or most important factor in a decision. PageUp says the feature does not use zip codes.
Watch out for. Its AI principles, ISO 42001 certificate and any model card are unpublished, and the Skills Matching help article sits behind a login. The audits run on test data and cover only Skills Matching, and its marketing calls the feature "free from bias", which the auditor's own disclaimers do not support.
Best for: employers that want a scoring method they can explain band by band.
3. SmartRecruiters - 63/100
SmartRecruiters, owned by SAP since September 2025, layers its Winston AI over an enterprise ATS: SmartAssistant and Winston Match score candidates against jobs, Winston Screen ranks them on screening answers, and Winston Interview is an agentic first-round interviewer.
Strengths. ISO 42001 certification, per-product AI impact assessments and DPIAs, a public AI whitepaper listing the identifiers removed from training data, and a July 2026 Warden AI LL 144 audit with monthly monitoring of SmartAssistant and Winston Match.
Watch out for. Winston Screen and Winston Interview, which rank and score candidates, have no public audit. Its AI Addendum tells customers its applications are not automated employment decision tools - despite the LL 144 audit - and still lists Colorado's repealed SB 205. The ISO certificate and impact assessments sit behind trust-centre requests.
4. Ashby - 59/100
Ashby is an all-in-one recruiting platform. Its opt-in AI includes AI-Assisted Application Review, whose Criteria Evaluation model marks each resume as meeting, not meeting or undecided on recruiter-set criteria, plus Talent Rediscovery, a notetaker, custom agents, and a voice AI Interviewer add-on announced in May 2026.
Strengths. Warden AI has audited Criteria Evaluation on 3.6 million real applications - rare among ATS audits, most of which use test data. Ashby's own Responsible AI statement says Application Review would be high-risk under the EU AI Act, and it makes a zero data retention commitment.
Watch out for. The AI Interviewer audit uses test profiles, the model cards are gated, and the Responsible AI statement still gives the old 2 August 2026 EU date. Ashby says its tooling is not an AEDT, though it publishes audits for customers who disagree.
5. Workday Recruiting AI - 59/100
Workday Recruiting AI now spans four surfaces: the native Candidate Skills Match, which Workday plans to retire; HiredScore grading; Paradox conversational screening and scheduling, acquired in October 2025; and the Talent Acquisition Agent released to early adopters in October 2026.
Strengths. ISO 42001 covering HCM, HiredScore and Paradox, a public administrator guide, detailed responsible-AI material, and joint-best customer documentation in the category.
Watch out for. Only HiredScore has public bias results; the other recruiting surfaces don't, and the per-feature AI Fact Sheets are not public. Workday is also the defendant in Mobley v. Workday, the most closely watched US case on vendor liability for AI hiring.
6-9. Dayforce, Pinpoint, Cornerstone, Hireology - 57/100
- Dayforce assigns external applicants AI letter grades with a weighted "report card". It announced ISO 42001 and a NIST AI RMF attestation in February 2026, runs an external AI Ethics Council, and has the strongest oversight score among ATSs. But it publishes no LL 144 audit: it suppresses grades for candidates in New York City instead, and its certificates and fairness audits are not public.
- Pinpoint layers eight AI features, including an AI Match Score and anonymised screening, on its ATS, and was among the first ATSs certified to ISO 42001. Its data governance disclosure is the best in the category and it describes its human-in-the-loop design as mandatory, but it has published no independent bias audit and no FRIA template.
- Cornerstone OnDemand scores applicants with a Candidate Skills Match built on skills inference, holds ISO 42001, and lets customers turn applicant scoring on only in chosen locations. Its only bias audit is a 2024 review of skill extraction, and its EU AI Act article still gives a mid-2026 timetable.
- Hireology sells AI Match and a voice AI Interview add-on to multi-location employers such as auto dealers. Its public Warden AI dashboard audits both monthly, giving it the second-best bias score among ATSs, but on test data. Its terms push deployer duties onto customers, and prescreen knockout rules can advance or eliminate candidates with no documented human review.
10. Rippling - 56/100
Rippling Recruiting screens applications against recruiter-defined criteria, summarises interviewer feedback with an LLM, transcribes video interviews and scores applications for fraud risk. It holds ISO 42001 and states plainly that under the EU AI Act it is a provider and its customers are deployers; its FRIA support ties Greenhouse for the category's best. The weakness is its Recruiting Bias Audit, which sits behind a registration wall, so the auditor, date and results cannot be checked.
Just behind, at 56, are Tribepad, whose Sidekick assistant scores applicants and whose claim that hires matched its top three 94% of the time comes with no method beyond a 15-case proof of concept, and Lever, whose Talent Fit audit covers an earlier version of the score than the one now in use. SAP SuccessFactors Recruiting (55) has strong corporate AI governance but no published bias results, and iCIMS (53) keeps its LL 144 audit summary in a gated portal.
What the AI ATS market is missing
- Audits of the right thing, on real data. 27 of the 52 score 30 or below on bias audit transparency. Among the leaders, Greenhouse, PageUp, Hireology and Tribepad audit on test data, and several audit one feature while shipping others that also score candidates.
- FRIA support. No ATS scores above 45. EU employers deploying ranking or screening features will be writing their own impact assessments.
- Current legal references. Greenhouse, SmartRecruiters and Lever's owner Employ still cite Colorado's repealed 2024 Act somewhere, and Ashby, iCIMS and Cornerstone give the EU AI Act's superseded 2026 timetable. The rules that apply are Colorado's SB 26-189 from 1 January 2027 and the EU's high-risk duties from 2 December 2027.
- AI interviewers arriving inside the ATS. Greenhouse, Ashby, SmartRecruiters and Hireology now sell AI interviewers, which score candidates more directly than matching does. Our AI interviewer ranking covers what they score and how they are audited.
How to choose an AI applicant tracking system
- List the AI features you will actually switch on. The duties attach to features, not the platform. Several vendors here let you enable scoring per office, location or job; use that.
- Ask which features the bias audit covers, and on what data. A matching audit says nothing about an AI interviewer or a screening score. See what makes an LL 144 audit defensible.
- Find out whether any AI output moves a candidate without a person. Knockout rules and score thresholds can do it even where the vendor says AI never decides.
- Check the vendor's legal dates. A statement written for Colorado's repealed Act or the EU's old timetable is out of date.
- Ask for the technical documentation. Our Annex IV walkthrough gives the questions, and the EU AI Act quick check tells you whether your use is high-risk.
For the wider market, see the best AI recruiting tools and the best AI sourcing tools. Every score above links to a full scorecard with the evidence behind each category, and any vendor can submit new evidence.
Frequently asked questions
What is an AI applicant tracking system?
An applicant tracking system with AI features built in: typically candidate matching or ranking against a job, resume screening, summaries and notetakers, and increasingly AI agents and AI interviewers. The ATS itself is a system of record; the AI features that score, rank or screen candidates are what AI hiring laws regulate.
What is the best AI applicant tracking system in 2026?
On HireAIScore's rubric, Greenhouse scores highest at 67/100, followed by PageUp (64) and SmartRecruiters (63). The scores rate the public evidence behind each vendor's AI features, not how well the ATS works, and no ATS reaches a C.
Is an AI-powered ATS covered by NYC Local Law 144?
The features that score or rank candidates can be, when an employer uses them as a substantial factor in a hiring decision for New York City roles. iCIMS, for example, names Candidate Ranking using Role Fit as the only automated employment decision tool in its platform. Pure tracking, scheduling and search features generally are not.
Do AI ATS bias audits use real applicant data?
Mostly not. Greenhouse, PageUp, Hireology and Tribepad publish Warden AI audits run on Warden's test data. Ashby is an exception: Warden audited its Criteria Evaluation on 3.6 million real applications, though its AI Interviewer audit uses test profiles.
How were these AI applicant tracking systems ranked?
Each is scored on HireAIScore's rubric v1.0 from public evidence: Article 11 technical documentation (20%), bias audit transparency (18%), FRIA support (15%), data governance disclosure (15%), human oversight design (12%), post-market monitoring (12%) and customer documentation (8%). The ranking is the overall score.