ZipRecruiter
ZipRecruiter is an online employment marketplace whose AI matching engine sends employers' job postings out, alerts matched job seekers and invites them to apply, and labels and ranks applicants (for example as 'Great Match') in the employer's candidate dashboard.
HireAIScore rates ZipRecruiter 40 out of 100 (grade F, Substantial gaps) under rubric v1.0, last reviewed September 25, 2026. ZipRecruiter ranks 66 of 86 vendors rated in Candidate sourcing AI, against a category average of 46. That score is drawn from 33 evidence items across 7 rubric criteria for ZipRecruiter — 28 with a cited source, 5 recording that nothing was located.
§ 01 - Company facts
- Legal name
- ZipRecruiter, Inc.
- Founded
- 2010
- Headquarters
- United States · US
- Pricing tier
- Mid
- Market side
- Employer-side (AEDT)
- Categories
- SourcingScreening
- Website
- ziprecruiter.com
These are company details, not findings. They are editable by a verified representative and carry no weight in the score, which is built only from the cited evidence below.
§ 02 - Score breakdown
§ Score breakdown
Category scoring
Weighted contribution shown to the right of each bar.
- 01
Article 11 Technical Documentation
Weight 20%40
+8.0 · category avg 48
- 02
Bias Audit Transparency
Weight 18%24
+4.3 · category avg 40
- 03
FRIA Support
Weight 15%27
+4.0 · category avg 31
- 04
Data Governance Disclosure
Weight 15%52
+7.8 · category avg 53
- 05
Human Oversight Design
Weight 12%55
+6.6 · category avg 56
- 06
Post-Market Monitoring
Weight 12%40
+4.8 · category avg 40
- 07
Customer Documentation
Weight 8%54
+4.3 · category avg 57
Category avg is the mean raw score on that criterion across the 86 Candidate sourcing AI vendors in scope of this rubric, this one included.
§ 03 - Strongest · weakest
Strongest category
Human Oversight Design
Raw score 55 · contributes 6.6 to total.
55 against a 56 category average
Weakest category
Bias Audit Transparency
Raw score 24 · contributes 4.3 to total.
24 against a 40 category average
§ 04 - Cited evidence
Download diligence record→§ Evidence
Cited per category
Every score is backed by at least one cited piece of evidence.
Evidence ledger
- Items
- 33
- Documentation
- 23
- Public statement
- 4
- Product feature
- 1
- Absence
- 5
- With a source URL
- 28 of 33
- Source hostnames
- 7
- Fewest items
- 4
- Bias Audit Transparency
These figures measure how thoroughly ZipRecruiter was reviewed, not how ZipRecruiter performed — an absence row, recording that nothing was located, is counted like any other item.
Article 11 Technical Documentation
6 items40
- DocumentationCaptured Sep 25, 2026
ZipRecruiter's Responsible AI page sets ten principles for the AI that matches employers and job seekers: transparency, fairness, accountability with performance logs, safety, security, legal compliance, human oversight, vendor review, generative-AI guardrails and privacy. It cites no framework and links to no system documentation.
- DocumentationCaptured Sep 25, 2026
An archived June 2018 product post is the only first-party source found that names the Great Match model's inputs: the applicant's skills, job titles, years of experience, recency of titles, location, resume length and application rate.
https://www.ziprecruiter.com/blog/candidate-calibration-feature/
- DocumentationCaptured Sep 25, 2026
The archived March 2021 launch post says the Great/Good/Fair/Not a Match labels come from AI trained on 'billions of employer and job seeker interactions' and show how likely an applicant is 'to be received favorably by a hiring manager', meaning the model predicts employer response.
https://www.ziprecruiter.com/blog/ziprecruiter-tells-you-how-strong-a-match-you-are-for-every-job/
- DocumentationCaptured Sep 25, 2026
The FY2025 10-K (signed 25 February 2026) describes the matching only at business level, as deep-learning NLP over clicks, applications, hiring signals and 'billions of user interactions', plus 'Matching that learns' from employer ratings, with no model, validation or performance detail.
https://www.sec.gov/Archives/edgar/data/1617553/000161755326000016/zip-20251231.htm
- DocumentationCaptured Sep 25, 2026
The only AI whitepaper found, the 2019 'AI Advantage' guide, is marketing copy saying the AI analyses resume data, searches and certifications and learns from candidate and employer signals; it has no methodology, validation or fairness content.
- AbsenceCaptured Sep 25, 2026
No model or system card, technical documentation pack, explainability statement, instructions for use or ISO/IEC 42001 certification was found on the ziprecruiter.global legal hub (Responsible AI, Security and Compliance, Terms, Privacy), the ziprecruiter.com employer and enterprise pages, the Employer Help Center or ziprecruiter-investors.com.
Bias Audit Transparency
4 items24
- AbsenceCaptured Sep 25, 2026
No NYC LL 144 bias-audit summary, independent audit (BABL AI, DCI, ORCAA, Warden AI, Holistic AI, Credo AI, ConductorAI) or academic audit of Great Match, match ratings, Invite to Apply or Sponsored Reach was found. A web search for a ZipRecruiter LL 144 audit returned only general material, the Employer Help Center returns no results for 'bias' or 'audit', and trust.warden-ai.com/ziprecruiter returns 404.
- Public statementCaptured Sep 25, 2026
The Responsible AI page says ZipRecruiter takes steps so its models 'do not directly use' protected-class characteristics and applies design, testing and monitoring principles against unintended bias, but it publishes no test results, selection rates or impact ratios.
- DocumentationCaptured Sep 25, 2026
The FY2025 10-K names New York City's regulation of automated employment decision tools, the Illinois and Colorado employment-AI laws, and the risk that models trained on biased data produce 'biased, discriminatory' outputs, yet it discloses no audit.
https://www.sec.gov/Archives/edgar/data/1617553/000161755326000016/zip-20251231.htm
- DocumentationCaptured Sep 25, 2026
The Global Terms of Use (effective 11 August 2026) have employers agree that the candidate-filtering functionality 'does not, and is not intended to, constitute, contribute to, or make a hiring recommendation or decision', which puts screening responsibility on employers instead of giving them audit data.
FRIA Support
4 items27
- AbsenceCaptured Sep 25, 2026
No EU AI Act Article 27 fundamental-rights impact assessment template, deployer guidance or AI Act statement was found on the ziprecruiter.global legal hub (Responsible AI, Security and Compliance, Privacy, Terms), the Employer Help Center or the investor site, even though ZipRecruiter runs EU job sites.
- DocumentationCaptured Sep 25, 2026
The Security and Compliance page's EU material covers GDPR only. It lists ZipRecruiter.de, .fr, .ie and .co.uk as in scope, names Article 27 GDPR representatives in the EU and UK, and offers a Data Processing Addendum where ZipRecruiter acts as processor, with nothing on AI Act deployer duties.
- DocumentationCaptured Sep 25, 2026
ZipRecruiter's live German site accepts employer sign-ups and tells job seekers its 'Matching-Technologie' delivers suitable job offers to their inbox, with no AI Act disclosure.
- DocumentationCaptured Sep 25, 2026
The 10-K's only treatment of the EU AI Act is a generic risk factor that expects most obligations to 'take effect in August 2026', with no compliance or deployer-support plan.
https://www.sec.gov/Archives/edgar/data/1617553/000161755326000016/zip-20251231.htm
Data Governance Disclosure
4 items52
- DocumentationCaptured Sep 25, 2026
The Global Privacy Policy effective 30 September 2026 adds a 'How AI is Used to Process Personal Data' section. It says identity, profile, technical and usage data may be used 'to help train, optimize, and power our AI systems', and that large language models and other generative AI help match job seekers and employers.
https://www.ziprecruiter.com/assets/static/pdf/legal/global-privacy-policy-en-2026-09-30.pdf
- Public statementCaptured Sep 25, 2026
The Responsible AI page states that the models do not directly use protected-class characteristics and that data minimisation and other privacy-enhancing techniques apply when data trains or powers AI, but it lists neither the excluded fields nor any proxy controls.
- DocumentationCaptured Sep 25, 2026
The Security and Compliance page documents AES-256 encryption at rest, TLS 1.2/1.3 in transit, AWS hosting, EU-U.S. Data Privacy Framework certification and a GDPR/CCPA programme, but its only SOC 2 Type 2 claim covers 2 April to 31 December 2022.
- DocumentationCaptured Sep 25, 2026
The 10-K says the matching learns from 'billions of user interactions', including 'how employers engage with every job seeker', so employer behaviour is itself training signal; it describes no bias controls on that data.
https://www.sec.gov/Archives/edgar/data/1617553/000161755326000016/zip-20251231.htm
Human Oversight Design
6 items55
- DocumentationCaptured Sep 25, 2026
The Terms of Use state that employers 'can view any applicant at any time', that ZipRecruiter 'makes no hiring recommendations or decisions about any applicant', and that its filtering is 'not a substitute for human discretion and review'.
- DocumentationCaptured Sep 25, 2026
Deal-breaker screening questions move applicants who miss basic qualifications to a 'Hidden' list the employer can still open, instead of rejecting them outright.
https://support.ziprecruiter.com/s/article/What-are-Screening-Questions
- DocumentationCaptured Sep 25, 2026
The QuickRate article says 'every time you give a thumbs up, our matching technology contacts similar job seekers and invites them to apply', so employer ratings drive automated outreach with no documented review of who gets invited.
https://support.ziprecruiter.com/s/article/How-do-I-use-QuickRate-to-review-my-candidates
- Product featureCaptured Sep 25, 2026
In the CareerPlug/ADP ATS integration, ZipRecruiter alone decides which applicants get the 'Great Match' label; employers can reject them to retrain the algorithm, but the label comes with no stated reasons.
- DocumentationCaptured Sep 25, 2026
Since its 2021 redesign, Invite to Apply shows employers each candidate's job titles, years of experience, education, veteran status, certifications and a match-strength rating, with no documented explanation of the rating for each candidate.
https://www.ziprecruiter.com/blog/ziprecruiters-invite-to-apply-feature/
- Public statementCaptured Sep 25, 2026
The Be Seen First release (22 January 2026) says applications carrying a personalised note are 'boosted to the top of the employer's list', so the dashboard order mixes applicant-intent signals with match.
Post-Market Monitoring
4 items40
- DocumentationCaptured Sep 25, 2026
The security.txt file (expires 1 July 2027) sends vulnerability reports to security@ziprecruiter.com and offers researchers entry to a private bug bounty programme.
- DocumentationCaptured Sep 25, 2026
The Security and Compliance page describes incident detection and response procedures, annual third-party penetration testing, and separate security@ and trustandsafety@ reporting addresses.
- DocumentationCaptured Sep 25, 2026
The Transparency Report page publishes half-yearly counts of government access requests from July 2020 to December 2025 and links an EU DSA report (published 31 March 2026, covering 2025). This is periodic public reporting, but none of it covers AI performance or fairness.
- AbsenceCaptured Sep 25, 2026
No public status page was found (status.ziprecruiter.com does not resolve and ziprecruiter.statuspage.io redirects to Statuspage's marketing site). There is also no AI model-update changelog, no AI-specific incident or contest channel and no published fairness-monitoring results; the Responsible AI page's claims of continuous monitoring and performance logs are not surfaced publicly.
Customer Documentation
5 items54
- DocumentationCaptured Sep 25, 2026
The public Employer Help Center explains rating, QuickRate, filtering, hiding and screening questions and says ratings teach the matching. A search for 'AI' returns one article, and none explains how Great Match labels or match ratings are calculated.
https://support.ziprecruiter.com/s/topic/0TO0f000000DncSGAS/rating-and-managing-candidates
- DocumentationCaptured Sep 25, 2026
The Enterprise FAQ says Sponsored Reach 'utilizes AI to analyze millions of data points including resume keywords, skills, experience, licenses, past searches and applications' and actively invites matched candidates to apply.
- Public statementCaptured Sep 25, 2026
The June 2026 Smart Outreach release describes AI-written, editable sequences of up to three messages sent automatically to Resume Database candidates the employer picks, and gives no fairness or compliance guidance.
- DocumentationCaptured Sep 25, 2026
A public Marketplace Services Privacy and Security Addendum covers US and Canadian end-user data for Marketplace clients but states it does not apply to ZipRecruiter.com services, which run under the Terms of Use only; neither document addresses AI.
https://www.ziprecruiter.global/en/marketplace-privacy-addendum
- AbsenceCaptured Sep 25, 2026
No deployer guidance for NYC LL 144, Illinois HB 3773, Colorado, California or the EU AI Act (candidate-notice templates, audit data, per-jurisdiction settings) was found on ziprecruiter.com, the ziprecruiter.global legal hub or the Employer Help Center.
§ 05 - Editorial notes
Company overview
ZipRecruiter, Inc. (NYSE: ZIP) runs one of the largest US online employment marketplaces. It was incorporated in Delaware in 2010 and is headquartered in Santa Monica, California. At the end of 2025 it had over 800 employees across the US, UK, Canada and Israel, 2025 revenue of $449.0M (down 5%), and 59,104 quarterly paid employers in Q4 2025, each bringing in about $1,900 of revenue that quarter. Employers buy Standard, Premium or Pro plans, or enterprise Sponsored Reach campaigns, to send jobs to 100+ boards and search a database of 56M+ resumes. The AI graded here is the employer-side matching engine. When a job goes live, it shows the employer a list of the 'best potential candidates' with a match rating so they can invite them in one click (Invite to Apply). It also alerts matched job seekers, labels applicants 'Great Match', and learns from employer ratings: every thumbs-up sends invitations to 'similar' job seekers. Two recent additions are Smart Outreach (June 2026), which sends AI-written sequences of up to three messages to database candidates the employer picks, and Be Seen First (January 2026), which moves applicants who add a note to the top of the employer's list. ZipRecruiter also runs UK, German, French and Irish sites and bought Breakroom in July 2024. Job-seeker tools (Phil, the ChatGPT and Claude apps) are not graded here.
Regulatory exposure
ZipRecruiter's FY2025 10-K lists New York City's AEDT law, the Illinois and Colorado employment-AI rules and the EU AI Act as risks, and warns that models trained on biased data can produce discriminatory output. The company has published nothing that addresses those risks. Its match model predicts how likely a hiring manager is to respond favourably, and it learns from employer ratings. That makes Great Match labels and match ratings on applicants the part of the product closest to NYC Local Law 144's definition of 'simplified output'. There is no bias audit, and the Terms of Use have employers agree that the filtering features do not 'constitute, contribute to, or make a hiring recommendation or decision'. That leaves every audit and candidate-notice duty with employers, who get no audit data to meet it, just as DCWP has committed to tougher enforcement after the December 2025 Comptroller audit. Targeting before anyone applies (Invite to Apply, Sponsored Reach, job alerts) falls outside LL 144, which covers applicants. It is squarely 'targeted job advertisements' under EU AI Act Annex III point 4(a), and ZipRecruiter runs German, French and Irish sites. Those duties now start on 2 December 2027, and ZipRecruiter has published nothing on the AI Act. Illinois HB 3773 (in force since 1 January 2026) requires notice of AI use and bans zip codes as proxies; the only disclosed input list includes location. Colorado's SB 26-189 disclosure regime starts on 1 January 2027. In its home state of California, the FEHA automated-decision rules (since 1 October 2025, with four-year record retention) apply now, and the CPPA ADMT rules apply from 1 January 2027.
Path to a higher score
The single biggest gain would come from an independent bias audit of Great Match labels, match ratings and Invite to Apply targeting, published as a downloadable summary with selection rates and impact ratios by sex, race/ethnicity and intersectional groups, and repeated every year. That would lift bias transparency from near the floor into the 55-75 band. Next, ZipRecruiter should publish a system card for the matching engine that cites a framework such as the NIST AI RMF or ISO/IEC 42001. It should cover the inputs (including how location, behavioural signals and employer-engagement signals are used), excluded fields and proxy testing, validation, and how employer ratings and Be Seen First change the order candidates appear in. Beyond that, it should show employers in the dashboard why a candidate earned a Great Match label, and give NYC, Illinois, Colorado, California and EU employers deployer guidance with candidate-notice language, audit data and inputs for an Article 27 FRIA. Finally, it should renew the SOC 2 claim (which covers only April to December 2022), add a public status page, and publish an AI model changelog.
§ Regulatory frame
What applies to candidate sourcing ai.
Ranking outputs are inside Annex III §4 when they influence hiring decisions. GDPR Article 22 (automated decision-making) and the GDPR rights around profiling apply independently for EU candidates.
§ Compare
Build any comparison→ZipRecruiter against its nearest-scoring peers.
In Candidate sourcing AI.
§ Others rated in Candidate sourcing AI
All sourcing vendors→Ranked 66 of 86 by weighted total under rubric v1.0. The ordering is arithmetic on the rubric and carries no view on which tool suits a given hiring process.
- 63Personio40F
- 64Pin40F
- 65Feenyx40F
- 66ZipRecruiterThis profile40F
- 67Paycor40F
- 68Arya by Leoforce38F
Conflicts of interest
No vendor pays for placement, scoring, or removal. Casework - the consulting firm that operates this directory - provides paid services to some vendors. Any active or recent (within 24 months) commercial relationship is disclosed on the affected vendor profile and the review is reassigned to an independent reviewer. See the full policy on About.
Casework has no commercial relationship with this vendor.