Feenyx
HireAIScore rates Feenyx 40 out of 100 (grade F, Substantial gaps) under rubric v1.0, last reviewed September 25, 2026. Feenyx ranks 65 of 82 vendors rated in Assessment AI, against a category average of 46. That score is drawn from 35 evidence items across 7 rubric criteria for Feenyx — 30 with a cited source, 5 recording that nothing was located.
Score breakdown
Feenyx, 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 50
42
+8.4
Bias Audit Transparency
Weight 18% · category avg 41
25
+4.5
FRIA Support
Weight 15% · category avg 31
25
+3.8
Data Governance Disclosure
Weight 15% · category avg 53
50
+7.5
Human Oversight Design
Weight 12% · category avg 55
58
+7.0
Post-Market Monitoring
Weight 12% · category avg 38
38
+4.6
Customer Documentation
Weight 8% · category avg 58
55
+4.4
Total weighted score
40.1(rounded to 40 for display)
Category avg is the mean raw score on that criterion across the 82 Assessment AI vendors in scope of this rubric, this one included.
Strongest criterion
Human Oversight Design
58raw / 100
Contributes 7.0 to the total. 58 against a 55 category average.
Weakest criterion
FRIA Support
25raw / 100
Contributes 3.8 to the total. 25 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
- 35
- Documentation
- 23
- Audit report
- 1
- Public statement
- 6
- Absence
- 5
- With a source URL
- 30 of 35
- Source hostnames
- 6
- Fewest items
- 2
- FRIA Support
These figures measure how thoroughly Feenyx was reviewed, not how Feenyx performed — an absence row, recording that nothing was located, is counted like any other item.
Article 11 Technical Documentation
4 items42
- DocumentationCaptured Sep 25, 2026
The short How Feenyx Uses AI page lists webcam face detection for proctoring (not facial recognition), NLP skill extraction and transcription, and generative-AI summaries and suggested interview questions, and says models are checked for bias and kept explainable without saying how.
- DocumentationCaptured Sep 25, 2026
A technical manual by I-O psychologist R. Chris Steilberg (TalentLift), marked as work in progress as of April 2018, covers the coding test's design, its scoring on solve time, memory use and accuracy with percentile norms, and a content-validity rationale; it does not cover today's generative-AI features.
- DocumentationCaptured Sep 25, 2026
A help-centre article dated June 2025 explains that coding answers are auto-scored against test cases with partial credit and that customers can set a pass threshold that filters out lower-scoring candidates.
https://help.feenyx.ai/en/articles/1953224-scoring-candidates
- AbsenceCaptured Sep 25, 2026
No model or system cards, instructions for use, explainability statement or ISO/IEC 42001 certification were found, and the AI-use page omits the marketed resume scoring, rubric auto-scoring, match ranking and reading-behaviour analysis (searched all 109 feenyx.ai sitemap pages and help.feenyx.ai).
Bias Audit Transparency
4 items25
- DocumentationCaptured Sep 25, 2026
The Compliance & bias page claims EEOC, OFCCP, ADA and Sarbanes-Oxley compliance and cites an external evaluation by TalentLift's R. Chris Steilberg, PhD, but reports no adverse-impact results.
- DocumentationCaptured Sep 25, 2026
The linked evaluation is a 2018 content-validity manual for the coding test with no impact-ratio or demographic subgroup analysis, so it is not a bias audit.
- Public statementCaptured Sep 25, 2026
The pricing page says analytics dashboards show pass rates, time-to-hire and diversity metrics, but no adverse-impact method or results are published.
- AbsenceCaptured Sep 25, 2026
No NYC Local Law 144 bias audit, impact ratios or independent audit was found: the terms bias audit, Local Law 144, AEDT and impact ratio appear nowhere in feenyx.ai's 109 sitemap pages or help.feenyx.ai, and trust.warden-ai.com/feenyx returns 404.
FRIA Support
2 items25
- AbsenceCaptured Sep 25, 2026
There is no EU AI Act material, FRIA template or deployer-obligation guidance: the AI Act, high-risk classification and fundamental-rights impact are not mentioned anywhere on feenyx.ai, its blog, its customer terms or help.feenyx.ai.
- DocumentationCaptured Sep 25, 2026
The only compliance guidance aimed at deployers is a Disability Guidelines page telling employers to run the ADA interactive process and offer reasonable accommodations in assessments.
Data Governance Disclosure
8 items50
- Audit reportCaptured Sep 25, 2026
INTERCERT certifies Filtered, Inc's ISO/IEC 27001:2022 information-security management system for its skills-assessment and candidate-interview SaaS (initial certification 15 May 2025, surveillance validity 14 May 2026, registration IC-IS-2505185).
- DocumentationCaptured Sep 25, 2026
An undated subprocessor list names Google Cloud, AWS, MongoDB, OpenAI (text summaries and recommendations), Veriff (identity verification), TalentIQ (public social profiles), IPStack (IP geolocation) and others, with their locations.
- DocumentationCaptured Sep 25, 2026
The security page covers daily backups, encryption in transit and at rest, Auth0 login, and logged employee access to customer data, with the customer's written consent obtained first where practical.
- DocumentationCaptured Sep 25, 2026
The reading-behaviour launch post rules out biometric identification and storage of raw face embeddings, but says original video and derived motion metrics are kept for audit and model improvement.
- DocumentationCaptured Sep 25, 2026
The privacy policy (effective 5 June 2023, still issued by Filtered, Inc) lists names, video, audio, work history and email but says nothing on AI processing, biometric data, model training or specific retention periods.
- Public statementCaptured Sep 25, 2026
The terms of service say candidates' videos, audio, resumes and metadata become Feenyx's property, which contradicts the homepage FAQ saying the customer and candidate own that data.
- DocumentationCaptured Sep 25, 2026
ID verification includes facial matching of a candidate's selfie to their government-ID photo, a biometric step that neither the AI-use page nor the privacy policy discloses.
- Public statementCaptured Sep 25, 2026
The 31 August 2026 Business Wire release announcing Recruiter.com's completed acquisition says Feenyx features will be integrated into Recruiter.com products and customers supported without interruption, but announces no data-policy changes.
https://finance.yahoo.com/technology/ai/articles/recruiter-com-acquires-feenyx-help-110000021.html
Human Oversight Design
7 items58
- DocumentationCaptured Sep 25, 2026
Auto-Score turns interview recordings into draft rubric scores; reviewers can override any score but must give a reason, which creates an auditable trail.
https://www.feenyx.ai/features/auto-score-for-live-interviews
- DocumentationCaptured Sep 25, 2026
Feenyx says it will not auto-reject candidates based on behaviour; fraud signals (tab use, code similarity, extra faces in snapshots, IP-location mismatch) go into a confidence report for human reviewers.
- DocumentationCaptured Sep 25, 2026
Reviewers make the final call on reading-behaviour flags, see the eye-path and head-pose metrics behind each flag, and can exclude segments, tune sensitivity and whitelist allowed behaviours.
- DocumentationCaptured Sep 25, 2026
The beta Interview Copilot suggests questions and scores answers from 0 to 10, but only starts analysing after the interviewer clicks to activate it for a specific question.
- Public statementCaptured Sep 25, 2026
The homepage FAQ says customers can turn off Feenyx's AI components and keep using the platform.
- Public statementCaptured Sep 25, 2026
The AI-scored video assessment page sells automated rubric scoring as removing human bias and speeding shortlisting, and describes no human review of those scores.
https://www.feenyx.ai/features/ai-scored-async-video-assessments
- Public statementCaptured Sep 25, 2026
Matchmaking scores and ranks candidates against every open role using assessment scores, behaviour and hiring outcomes, with no review step described.
Post-Market Monitoring
5 items38
- DocumentationCaptured Sep 25, 2026
Feenyx runs an UptimeRobot-hosted status page with email subscriptions and an update history, but the marketing site doesn't link to it.
- DocumentationCaptured Sep 25, 2026
The status page's data covers 9 services (API gateway, compiler, recording, data and others) with 90-day uptime, including a roughly five-day Workspace Proxy Service outage from 3 September 2026, and no monitor carries incident comments.
https://status.feenyx.ai/api/getMonitorList/873VpSXxgP?page=1
- DocumentationCaptured Sep 25, 2026
The security page gives only a support email for security concerns and a general promise to keep customers informed of incidents; there is no vulnerability-disclosure policy, and security.txt returns 404.
- DocumentationCaptured Sep 25, 2026
The Auto-Score launch post suggests customers track override rates, inter-rater alignment and pass-through by criterion, but Feenyx publishes no monitoring of its own models' performance or drift.
https://www.feenyx.ai/blog/introducing-auto-score-for-feenyx-interviews
- AbsenceCaptured Sep 25, 2026
No AI model changelog, AI incident-reporting channel or public fairness-monitoring dashboard was found on feenyx.ai, its blog (product news only) or help.feenyx.ai.
Customer Documentation
5 items55
- DocumentationCaptured Sep 25, 2026
A public help centre holds about 50 articles for recruiters and candidates in five collections, covering interview setup, scoring, confidence analysis, ATS integrations and live rooms.
- DocumentationCaptured Sep 25, 2026
Feature pages such as reading-behaviour detection include deployer FAQs on false positives, stored data and accommodations for candidates.
- DocumentationCaptured Sep 25, 2026
The Disability Guidelines page gives employers ADA reasonable-accommodation guidance for assessments.
- DocumentationCaptured Sep 25, 2026
The public standard customer terms put legal compliance and indemnity for candidate claims on the customer, bar customers from publicly disseminating performance information, allow client data to be deleted 30 days after termination and reference no DPA.
- AbsenceCaptured Sep 25, 2026
No DPA, candidate AI-notice template, or deployer guidance for NYC LL 144, the EU AI Act, Illinois or Colorado was found on feenyx.ai (109 sitemap pages) or help.feenyx.ai.
Editorial notes
Company overview
Feenyx is the rebranded Filtered.ai, operated by Filtered, Inc. of Boston, Massachusetts. It began in 2016 as an internal project at a consultancy and spun out in 2019. The platform combines asynchronous skills assessments (coding, simulations, and AI-scored video and written answers), AI-assisted live interviews with transcription and rubric auto-scoring, an AI agent for resume review and candidate comparison, and 'matchmaking' that scores and ranks talent-pool candidates. Fraud and AI-use detection runs throughout: webcam face counts, reading-behaviour analysis, tab tracking and selfie-to-ID matching. Published pricing starts at $25 per user a month for live interviews and $95 a month for assessments, and the security page cites ISO/IEC 27001. On 31 August 2026 Recruiter.com announced it had completed the acquisition. The company now operates as "Feenyx a Recruiter.com Company", and its capabilities will be integrated into Recruiter.com products. As of 25 September 2026 its privacy policy (effective June 2023) and customer terms still name Filtered, Inc., and no post-acquisition policy update was found.
Regulatory exposure
Feenyx auto-scores coding, video and written answers, scores and ranks resumes and matches, and lets customers set pass thresholds that screen candidates out. NYC employers using it to substantially assist hiring decisions would therefore likely be using an AEDT under Local Law 144. Feenyx publishes no bias audit, and its standard terms bar customers from publicly disseminating information about the service's performance, which would need clarifying before an audit summary is posted. Under the EU AI Act, recruitment and selection is high-risk (Annex III point 4). The Digital Omnibus deferred those obligations to 2 December 2027, but the ban on emotion recognition in the workplace has applied since 2 February 2025. Feenyx markets analysis of candidates' tone, pauses and glances, so deployers need assurance that none of it infers emotion; Feenyx publishes nothing on the Act. In Illinois, AI analysis of recorded video interviews brings the AI Video Interview Act into play, HB 3773 notice duties apply, and selfie-to-ID facial matching raises biometric questions the privacy policy doesn't address. Colorado's SB 26-189 disclosure regime starts on 1 January 2027. California's FEHA automated-decision rules already apply, including four-year record retention, while Feenyx's terms let it delete client data 30 days after termination; CPPA ADMT rules follow on 1 January 2027.
Path to a higher score
The biggest gain would come from commissioning an independent LL 144-style bias audit of the AI-scored assessments, resume scoring and match ranking, publishing its summary with impact ratios by sex, race/ethnicity and intersectional categories, and repeating it every year. Feenyx should also expand its How Feenyx Uses AI page to cover every scoring and ranking feature, and replace the 2018 coding-test manual with current system documentation and instructions for use. EU AI Act deployer guidance and a FRIA template should follow. The 2023 privacy policy needs a post-acquisition rewrite that covers AI, biometric and model-improvement processing, retention periods and the new owner. Feenyx should also reconcile its conflicting data-ownership statements, publish a DPA, and carve compliance disclosures out of the ban on publishing performance information. Finally, it should link its status page with incident notes, publish a model changelog and add a vulnerability-disclosure channel.
Regulatory frame
What applies to assessment AI.
Assessment is the most-regulated surface in AI hiring. NYC LL 144, EEOC and OFCCP frameworks, Illinois HB 3773, and EU AI Act Annex III §4 all apply, with overlap on bias-audit obligations.
Compare
Feenyx against its nearest-scoring peers.
The closest scores in Assessment AI, side by side.
- Side-by-sideFeenyx vs eSkill41F
- Side-by-sideFeenyx vs Willo41F
- Side-by-sideFeenyx vs Pillar40F
- Side-by-sideFeenyx vs Cappfinity40F
Ranking
Others rated in Assessment AI
Ranked 65 of 82 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.
- #62eSkill41F
- #63Willo41F
- #64Pillar40F
- #65FeenyxThis profile40F
- #66Cappfinity40F
- #67Adaface39F
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.