Crosschq
HireAIScore rates Crosschq 47 out of 100 (grade F, Substantial gaps) under rubric v1.0, last reviewed September 25, 2026. Crosschq ranks 67 of 160 vendors rated in Screening AI, against a category average of 46. That score is drawn from 43 evidence items across 7 rubric criteria for Crosschq — 38 with a cited source, 5 recording that nothing was located.
Score breakdown
Crosschq, criterion by criterion
Each rubric criterion is scored out of 100 on the cited evidence, then weighted into the total.
Weighted contribution shown to the right of each bar.
Article 11 Technical Documentation
Weight 20% · category avg 49
50
+10.0
Bias Audit Transparency
Weight 18% · category avg 41
38
+6.8
FRIA Support
Weight 15% · category avg 31
30
+4.5
Data Governance Disclosure
Weight 15% · category avg 53
58
+8.7
Human Oversight Design
Weight 12% · category avg 56
62
+7.4
Post-Market Monitoring
Weight 12% · category avg 40
40
+4.8
Customer Documentation
Weight 8% · category avg 58
62
+5.0
Total weighted score
47.2(rounded to 47 for display)
Category avg is the mean raw score on that criterion across the 160 Screening AI vendors in scope of this rubric, this one included.
Strongest criterion
Customer Documentation
62raw / 100
Contributes 5.0 to the total. 62 against a 58 category average.
Weakest criterion
FRIA Support
30raw / 100
Contributes 4.5 to the total. 30 against a 31 category average.
Cited evidence
The evidence behind the score
Every score is backed by at least one cited piece of evidence. Open a criterion to see what was found, when, and where.
Download diligence recordEvidence ledger
- Items
- 43
- Documentation
- 28
- Public statement
- 10
- Absence
- 5
- With a source URL
- 38 of 43
- Source hostnames
- 7
- Fewest items
- 3
- FRIA Support
These figures measure how thoroughly Crosschq was reviewed, not how Crosschq performed — an absence row, recording that nothing was located, is counted like any other item.
Article 11 Technical Documentation
7 items50
- DocumentationCaptured Sep 25, 2026
A Responsible & Ethical AI page sets out five principles (fair, transparent, tested/validated/monitored, reliable and safe, strict data privacy) without referencing NIST AI RMF, ISO 42001 or the EU AI Act.
- DocumentationCaptured Sep 25, 2026
The Quality of Hire model page describes a four-input model designed by Crosschq's Scientist in Residence and claims statistical-parity fairness metrics, but gives no methodology behind its "Explainable Algorithms" heading.
- DocumentationCaptured Sep 25, 2026
The AI Interview Agent page discloses the scoring architecture: multiple independent AI graders evaluate each question, and a supervisor model aggregates them into the final score.
- DocumentationCaptured Sep 25, 2026
A help article explains that Resume Screen uses LLM parsing and embeddings plus machine learning to produce a match score weighted across work experience, skills and education.
- DocumentationCaptured Sep 25, 2026
A public 14-page Traitify Big Five manual reports internal consistency (alpha > 0.70), test-retest reliability (r > 0.65) and construct validity against BFI/IPIP-NEO (n=942), deferring detailed subgroup results to a non-public full-length manual.
https://cdn.traitify.com/creative/downloads/Traitify_BigFive_Manual.pdf
- DocumentationCaptured Sep 25, 2026
The Drata-powered Trust Center lists an "AI Security and Ethical Usage Policy" among 10 policies, all available only through Request Access.
- AbsenceCaptured Sep 25, 2026
No model or system cards, ISO 42001 certification, Annex IV-style technical file, or accuracy/error metrics for the scoring or fraud models were found in the crosschq.com sitemap, the 204-article support.crosschq.com knowledge base, or the Trust Center.
Bias Audit Transparency
7 items38
- Public statementCaptured Sep 25, 2026
Traitify, acquired by Crosschq on March 31, 2026, says its most recent NYC LL 144 bias audit was completed by Holistic AI on April 21, 2025 with no selection-rate differences by gender or ethnicity across 400,000+ candidates, but the linked audit summaries now redirect to a login page.
- Public statementCaptured Sep 25, 2026
Crosschq claims "Multiple 3rd party experts auditing and validating models," naming only its own Scientist in Residence Steve Hunt and the investment firm Rocketship VC, and publishes no audit report.
- Public statementCaptured Sep 25, 2026
Crosschq's own 2021 funding release lists rocketship.vc as a returning investor, so the firm named as a third-party validator is not independent.
https://www.crosschq.com/press/2021/09/21/crosschq-raises-30m
- Public statementCaptured Sep 25, 2026
Crosschq markets 360 reference checks as validating skills and culture match "all without bias," a claim no published audit supports.
- Public statementCaptured Sep 25, 2026
AI Candidate Match produces "a ranked score" for applicants and is marketed as removing unconscious bias, with no bias testing or audit disclosed.
- DocumentationCaptured Sep 25, 2026
Admins can enable optional end-of-interview demographic questions "to support bias analysis," which lets customers analyze fairness themselves but is not an audit.
https://support.crosschq.com/knowledge/managing-crosschq-interview-settings
- AbsenceCaptured Sep 25, 2026
No LL 144 bias audit or impact-ratio summary exists for Crosschq's own interview, resume-screen or candidate-match models; searched crosschq.com (site searches for "Local Law 144", "bias audit", "AEDT"), trust.crosschq.com, support.crosschq.com, and trust.warden-ai.com/crosschq (404).
FRIA Support
3 items30
- DocumentationCaptured Sep 25, 2026
The DPA (last updated July 25, 2025) commits Crosschq to "reasonable assistance" with customers' GDPR data protection impact assessments on request, with no FRIA or EU AI Act content.
- DocumentationCaptured Sep 25, 2026
A trust package (SOC 2 report, penetration test, 10 security and data policies) is available only on request through the Drata Trust Center.
- AbsenceCaptured Sep 25, 2026
No FRIA template, EU AI Act deployer guidance or Annex III risk classification was found on crosschq.com (a site search for "EU AI Act" returned nothing), in the Trust Center, or in the support.crosschq.com knowledge base.
Data Governance Disclosure
7 items58
- DocumentationCaptured Sep 25, 2026
A help article gives an exhaustive list of what is redacted from AI-interview transcripts before scoring, which cannot be switched off: gender, age, race/ethnicity, religion, nationality, third-party names and direct identifiers. It also states that disability, neurodivergence, pregnancy and sexual orientation are not currently redacted.
https://support.crosschq.com/knowledge/transcript-redaction-in-crosschq-interview
- DocumentationCaptured Sep 25, 2026
A dated subprocessor list (Nov 19, 2025) names OpenAI, Copyleaks (AI-text detection), SEON, APIVoid and Fingerprint (fraud signals), and Namsor (optional candidate data enrichment).
- DocumentationCaptured Sep 25, 2026
Crosschq discloses an optional paid add-on that uses "inferential technology" (name morphology plus proprietary ML) to identify race and gender across the whole applicant pool for diversity reporting.
https://support.crosschq.com/knowledge/enriched-diversity-data
- DocumentationCaptured Sep 25, 2026
For ID verification, ID.me holds the identity documents and Crosschq stores only the verified result (name, email, birth date, phone, postal code); candidates can decline to verify.
https://support.crosschq.com/knowledge/crosschq-identity-verification
- DocumentationCaptured Sep 25, 2026
Crosschq states that personal data is not retained or used by any sub-processor to train their AI models, and commits to collecting only necessary data.
- DocumentationCaptured Sep 25, 2026
Crosschq (Danville, CA) is an active self-certified participant in the EU-U.S. DPF, its UK Extension and the Swiss-U.S. DPF since 09/12/2025, covering non-HR data, with Woofound Inc. dba Traitify listed as a covered entity.
- DocumentationCaptured Sep 25, 2026
The privacy policy (last updated Aug 27, 2025) discloses only that deleted data may persist in backups for up to 90 days, with no retention schedule for interview videos and no biometric-data disclosures.
Human Oversight Design
7 items62
- DocumentationCaptured Sep 25, 2026
The Fraud Analysis guide says the tool "does not make a hiring decision automatically" and that a high risk score "does not confirm that fraud occurred." It shows per-signal-group confidence and tells teams to verify identity through another channel, noting benign causes such as poor lighting.
https://support.crosschq.com/knowledge/fraud-analysis-in-crosschq-interview
- DocumentationCaptured Sep 25, 2026
Admins can switch off the AI recommendation score or set its Not Recommended/Neutral/Recommended thresholds, set video to Optional/Required/Disabled, require candidates to confirm their transcript, and enable or disable fraud detection.
https://support.crosschq.com/knowledge/managing-crosschq-interview-settings
- DocumentationCaptured Sep 25, 2026
The interview report explains scores in the interface with per-competency rubric anchor text, an AI summary with strengths and considerations, and severity-tagged alerts with explanatory tooltips.
https://support.crosschq.com/knowledge/reading-the-interview-report
- DocumentationCaptured Sep 25, 2026
Crosschq describes AI Suggestions as "a starting point, not a final verdict" and Fraud Analysis as "a tool to prompt further review, not an automated disqualification system."
https://support.crosschq.com/knowledge/interview-intelligence
- Public statementCaptured Sep 25, 2026
Crosschq states "Final hiring decisions are always made by humans — never AI" and that candidates can review transcripts, retake, or opt out entirely.
- Public statementCaptured Sep 25, 2026
Marketing copy says the AI Interview agent can "conduct initial interviews and screen out the noise," which sits in tension with the human-decision claim.
- DocumentationCaptured Sep 25, 2026
Resume Screen shows each applicant's match score as a five-star rating with an education/experience/skills breakdown, sorts applicants highest to lowest, and allows bulk advance or disposition.
Post-Market Monitoring
5 items40
- DocumentationCaptured Sep 25, 2026
An UptimeRobot-hosted "Crosschq US" status page, not linked from crosschq.com, runs 10 automated HTTP monitors (including Crosschq 360, Interview, Identity Verification and Partner API) but showed no incident posts or postmortems in the prior 30 days.
- DocumentationCaptured Sep 25, 2026
The Drata Trust Center lists continuous-monitoring controls (quarterly vulnerability scans, annual penetration tests, servers and databases monitored and alarmed), with a gated Incident Response Plan and Logging and Monitoring Policy.
- Public statementCaptured Sep 25, 2026
Crosschq claims "continuous active monitoring, testing and training" of its models and a customer feedback mechanism for reporting fairness and bias concerns, but publishes no monitoring results.
- Public statementCaptured Sep 25, 2026
Crosschq commits to "swift and transparent action" when its AI fails, but describes no public incident channel.
- AbsenceCaptured Sep 25, 2026
No security.txt (/.well-known/security.txt returns 404), vulnerability disclosure policy, dedicated security contact, AI-incident channel, or model-update changelog (/release-notes and /changelog return 404) was found, and crosschq.statuspage.io does not exist.
Customer Documentation
7 items62
- DocumentationCaptured Sep 25, 2026
A public DPA includes SCCs, sub-processor change notices with a 30-day objection window, annual audit rights, breach notification without undue delay, and DPIA assistance.
- DocumentationCaptured Sep 25, 2026
A published deletion process lets candidates and references request deletion via support@crosschq.com, completed within one week, with reference data anonymized in 360 reports.
https://support.crosschq.com/knowledge/candidate-or-reference-initiated-deletion-requests
- DocumentationCaptured Sep 25, 2026
Deployer guidance explains 360 fraud-alert severities and the LinkedIn-matching method (a 0.75 confidence threshold, with model reasoning saved for audit), and states that alerts prompt review rather than decide.
https://support.crosschq.com/knowledge/understanding-fraud-risk-in-crosschq-360
- DocumentationCaptured Sep 25, 2026
A public product page describes the layered fraud checks (ID-to-selfie matching, AI text forensics, deepfake/voice-clone detection, reference validation) and the resulting fraud score, without accuracy data.
- DocumentationCaptured Sep 25, 2026
The candidate terms (last updated Dec 4, 2025) contain no disclosure of, or consent to, AI analysis, video/audio recording or biometric processing, which Illinois AIVIA/BIPA notices would require.
- Public statementCaptured Sep 25, 2026
A Crosschq blog post summarizes NYC's AEDT law, the EU AI Act, Illinois and Colorado rules but says nothing about Crosschq's own compliance posture.
- AbsenceCaptured Sep 25, 2026
No NYC LL 144, Illinois AI Video Interview Act/BIPA, Colorado AI Act or EU AI Act deployer guidance was found on crosschq.com, support.crosschq.com (204-article sitemap reviewed) or the Trust Center.
Editorial notes
Company overview
Crosschq, Inc. was founded in 2018. Its press releases are datelined San Francisco, and its legal documents give an address in Danville, California. It sells a "hiring intelligence" suite: Crosschq 360 digital reference checks, and an AI Interview Agent that runs voice or video screening interviews and scores each answer with a panel of AI graders. It also sells Resume Screen and AI Candidate Match ranking, ApplicantX and embedded fraud detection, ID.me identity verification, and TalentWall recruiting analytics. On March 31, 2026 it acquired Traitify, which was previously part of Paradox, adding Big Five personality and cognitive assessments. Crosschq says its models are trained on more than 200 million hiring outcomes and that it integrates with over 40 ATS and HRIS partners.
Regulatory exposure
Several modules score or rank candidates: Interview Agent scores with Recommended/Not Recommended labels, a five-star Resume Screen match score, and AI Candidate Match's "ranked score." Crosschq also markets the agent as able to "screen out the noise." That places these modules in NYC LL 144 AEDT territory and in the EU AI Act's Annex III employment high-risk class, and Colorado's AI Act also treats hiring as a consequential decision. The only independent audit on record is Traitify's April 2025 Holistic AI LL 144 audit. It was commissioned under Paradox before the acquisition, is now more than a year old, and its linked summaries sit behind a login. Crosschq's own "3rd party" model validators are its Scientist in Residence and rocketship.vc, which is a Crosschq investor. The fraud tooling analyzes interview video (a still frame for deepfake checks, while marketing also claims eye-movement and response-timing analysis) and ID.me selfie matching. Yet Crosschq publishes no Illinois BIPA or AI Video Interview Act notice, consent or retention terms, and no fraud-model accuracy or false-positive rates. Its fraud signals include VoIP phone lines, LinkedIn presence and follower counts, and sanctioned-country location, and their disparate impact is unpublished. An optional add-on infers applicants' race and gender from their names for diversity reporting.
Path to a higher score
Commission and publicly host an independent LL 144-style bias audit that covers the AI Interview Agent, Resume Screen and AI Candidate Match, not only Traitify, and refresh the lapsed Traitify audit so it is publicly readable. Publish accuracy and false-positive data for the deepfake and fraud models, plus a BIPA / Illinois AI Video Interview Act statement on notice, consent and retention. Release model cards or instructions for use and an EU AI Act deployer / FRIA guide. Either back the "without bias" and "bias is eradicated" claims with audit evidence or remove them. Add a security.txt and a vulnerability-disclosure channel, link the status page and use it for incident posts, and publish a model-change log.
Regulatory frame
What applies to screening AI.
Treated as high-risk under Annex III §4 when the screening output is decisive. EU AI Act Article 13 transparency obligations apply; deployers must give candidates a way to contest.
Compare
Crosschq against its nearest-scoring peers.
The closest scores in Screening AI, side by side.
- Side-by-sideCrosschq vs isolved47F
- Side-by-sideCrosschq vs Covey47F
- Side-by-sideCrosschq vs Teamtailor47F
- Side-by-sideCrosschq vs Interviewer.AI47F
Ranking
Others rated in Screening AI
Ranked 67 of 160 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.
- #64Thomas International48F
- #65Jobylon48F
- #66isolved47F
- #67CrosschqThis profile47F
- #68Covey47F
- #69Teamtailor47F
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.