# Harri: HireAIScore fact sheet

URL: https://www.hireaiscore.com/vendors/harri
Fact sheet: https://www.hireaiscore.com/vendors/harri/facts.md
Last reviewed: 2026-08-02
Rubric: v1.0 (https://www.hireaiscore.com/methodology)

> HireAIScore rates Harri 41 out of 100 (grade F, Substantial gaps) under rubric v1.0, last reviewed August 2, 2026. Harri ranks 60 of 77 vendors rated in ATS-integrated AI, against a category average of 47. That score is drawn from 27 evidence items across 7 rubric criteria - 19 with a cited source, 8 recording that nothing was located.

This file is the text of the profile at the URL above, for AI assistants and other software. Scores are an editorial assessment of public evidence against a published rubric. They are not legal advice, a certification, or a statement that a product is or is not compliant.

## Score

- Overall: 41/100, grade F (Substantial gaps)
- Rank: 60 of 77 in ATS-integrated AI (category average 47)

## Criterion scores and evidence

### Article 11 Technical Documentation: 35/100 (weight 20%)

- Documentation (captured 2026-08-02): Harri's only dedicated AI page describes Salli and its Talent Gap Alert, Compliance Warning and Overtime Cost Alert agents in marketing terms, with no model description, training methodology, performance metrics, limitations or explainability statement. Source: https://harri.com/agentic-ai/
- Documentation (captured 2026-08-02): Addendum A (Artificial Intelligence Terms Addendum) of the public Master SaaS Agreement is the nearest thing to instructions for use, framing AI Outputs as inputs to human judgment while disclaiming that they will be 'accurate, complete, reliable, error-free'. Source: https://harri.com/msa/
- Searched, not found (captured 2026-08-02): No system card, model card, technical whitepaper, algorithmic transparency statement or ISO/IEC 42001 certification was found anywhere in harri.com's sitemap (which lists only marketing, legal and industry pages), on the agentic-ai page, in the resources hub, or in the privacy notice.

### Bias Audit Transparency: 22/100 (weight 18%)

- Searched, not found (captured 2026-08-02): No NYC Local Law 144 bias audit, independent auditor engagement (BABL AI, DCI, ORCAA, Warden AI, Holistic AI, Credo AI, ConductorAI or similar), academic audit, or adverse-impact/selection-rate disclosure was found; searches of the harri.com sitemap and of the resources.harri.com content index for 'AEDT', 'bias audit' and 'Local Law 144' returned no results.
- Documentation (captured 2026-08-02): Harri's page titled 'compliance' covers only wage-and-hour and scheduling law (breaks, overtime, minor employment, fair workweek) and contains no reference to bias audits, EEOC adverse impact, AEDT rules or algorithmic fairness. Source: https://harri.com/compliance/
- Documentation (captured 2026-08-02): The AI Terms Addendum's only bias-adjacent control is a prohibition on customers supplying data that could identify race, sex/gender, gender identity, national origin, sexual orientation, age, religion, disability, genetic information, medical condition, pregnancy or marital status - an input restriction, not an outcome audit. Source: https://harri.com/msa/
- Searched, not found (captured 2026-08-02): Harri's resources hub carries substantial DEI and 'unconscious bias' content aimed at human recruiters, but a content-index search for algorithmic or third-party AI audits returned nothing, confirming no published audit of Carri's own screening.

### FRIA Support: 44/100 (weight 15%)

- Documentation (captured 2026-08-02): Harri's public UK master agreement AI Terms Addendum explicitly requires customers to conduct 'fundamental rights impact assessments (including as contemplated by Article 27 of the EU AI Act)' and bars use constituting 'a prohibited artificial intelligence practice under Article 5 of the EU AI Act'. Source: https://harri.com/uk-msa/
- Documentation (captured 2026-08-02): The same addendum sets out Article 26-style deployer transparency duties, requiring the customer to inform 'affected individuals, including workers and workers' representatives, prior to their being subject to the AI Services' and to provide 'all required transparency disclosures regarding the use of the AI Services in employment-related decision-making'. Source: https://harri.com/uk-msa/
- Searched, not found (captured 2026-08-02): No FRIA template, questionnaire, worked example or Article 27 deployer guide is published anywhere on harri.com or resources.harri.com; the obligation is asserted contractually and left entirely to the customer to discharge, and the US-form MSA contains no EU AI Act references at all.

### Data Governance Disclosure: 52/100 (weight 15%)

- Documentation (captured 2026-08-02): The AI Terms Addendum discloses that Harri 'may use De-identified Data...for the sole purpose of developing, training, and improving Harri's artificial intelligence models, algorithms, and related services', and sets an explicit exclusion list barring customers from submitting protected-attribute data. Source: https://harri.com/msa/
- Documentation (captured 2026-08-02): Harri publishes a full UK Data Sharing Agreement incorporating the ICO's International Data Transfer Agreement (Version A1.0, in force 21 March 2022) with Harri as processor, plus a general sub-processor authorisation and objection process. Source: https://harri.com/uk-idta/
- Documentation (captured 2026-08-02): Site-wide footer badges assert AICPA SOC and 'ISO 27001 Certified' via auditor A-LIGN, but no report, scope statement, certificate number or NDA request process is offered anywhere on the site. Source: https://harri.com/about-us/
- Documentation (captured 2026-08-02): The privacy notice (last updated 1 February 2026, Harri US LLC) enumerates collected categories including resumes, job history, references, education, photographs, IP and device identifiers, and addresses CCPA/CPRA and Nevada rights - but contains no GDPR section and no explanation of how 'Agentic AI' processes personal information or makes automated decisions. Source: https://harri.com/privacy/
- Searched, not found (captured 2026-08-02): There is no security, trust or subprocessor-list page: harri.com/security returns HTTP 410, no trust centre appears in the sitemap, the Zendesk help centre at support.harri.com returns 403 to the public, and a resources-index search for 'ISO 27001' returned no results.

### Human Oversight Design: 52/100 (weight 12%)

- Documentation (captured 2026-08-02): Addendum A contractually mandates human oversight: customers 'will not rely solely on the AI Services or AI Outputs for any employment, scheduling, hiring, or other personnel decisions', reviewers must exercise 'independent judgment and due diligence', be given 'a meaningful decision-making role', and be trained 'in recognizing automation bias or algorithmic aversion'. Source: https://harri.com/msa/
- Documentation (captured 2026-08-02): The UK agreement goes further, barring use that would cause the AI Services to 'autonomously make, or materially influence, employment, scheduling, hiring, task allocation, or other personnel-related decisions without meaningful human oversight and independent judgment'. Source: https://harri.com/uk-msa/
- Public statement (captured 2026-08-02): Harri describes an accept/decline control on Salli's recommendations - managers review flagged issues before a schedule is published and 'the human remains in control'- but the post documents no override logging, confidence scores or output-level explanations. Source: https://resources.harri.com/blog/the-future-is-human-agentic-ai-for-hospitality/
- Searched, not found (captured 2026-08-02): No public documentation of in-product explainability for Carri's screening outputs, reviewer audit logs, adverse-action workflows, or per-jurisdiction configuration toggles; the AI Addendum refers only vaguely to 'electronic log data to the extent such data is made available to Customer', and the help centre is not publicly accessible.

### Post-Market Monitoring: 40/100 (weight 12%)

- Documentation (captured 2026-08-02): Harri operates a public status page covering seven components (Login & SSO, Talent Acquisition, Workforce Management, CoreHR, Carri, TeamHub, HarriIQ) with 90-day incident history - including a roughly 7.5-hour degradation on 21 July 2026 - and email, RSS, JSON, webhook and Slack subscriptions. Source: https://status.harri.com/
- Documentation (captured 2026-08-02): The privacy notice publishes a data-protection contact (DPO@harri.com and a toll-free number) for privacy enquiries and rights requests, which is the only named governance contact channel on the site. Source: https://harri.com/privacy/
- Searched, not found (captured 2026-08-02): No security.txt (harri.com/.well-known/security.txt returns HTTP 410), no vulnerability disclosure or bug bounty programme, no AI incident-reporting channel, no model-update changelog, and no continuous fairness or drift monitoring dashboard was found on harri.com, status.harri.com or resources.harri.com.

### Customer Documentation: 60/100 (weight 8%)

- Documentation (captured 2026-08-02): A full talent-acquisition product page covering Talent Scout, advanced talent pools, targeted campaigns, the conversational hiring assistant, applicant tracking and a custom career portal, positioning Harri as an AI hiring platform for frontline teams. Source: https://harri.com/pre-hire/
- Documentation (captured 2026-08-02): Harri publishes its entire Master SaaS Subscription Agreement openly, including Addendum A (AI Terms) and Addendum B (Data Processing Addendum), giving prospective deployers pre-signature visibility into AI and data-processing obligations. Source: https://harri.com/msa/
- Documentation (captured 2026-08-02): A substantial resources hub spanning blog, guides and ebooks, case studies, news and events/webinars, with recurring quarterly compliance advisories and deep NYC Fair Workweek, California PAGA and UK Employment Rights Act coverage. Source: https://resources.harri.com/
- Public statement (captured 2026-08-02): A blog post frames Harri's approach around 'transparency, security, and responsible technology use' and advises operators to be open about what AI systems are used and how they work, and to ensure technologies are 'fair and unbiased, and rigorously audited'- as general principles, not product commitments. Source: https://resources.harri.com/blog/balancing-ai-innovation-and-employee-trust-in-hospitality/
- Searched, not found (captured 2026-08-02): No NYC Local Law 144 notice templates, AEDT guidance, EU AI Act deployer guide, AI FAQ or candidate-facing AI explanation exists; harri.com/terms/ pushes AI and automated-decision-system compliance entirely onto the user, and the Zendesk help centre at support.harri.com returns 403 to non-customers.

## Company facts

- Website: https://harri.com
- Legal name: Harri US LLC
- Founded: 2012
- Headquarters: United States (US)
- Pricing tier: Enterprise
- Market side: Employer-side (AEDT)
- Categories: ATS-integrated AI, Screening AI
- Description: Harri is a frontline workforce platform for hospitality, grocery and care employers whose pre-hire suite pairs a career portal and applicant tracking system with Carri, a conversational AI assistant that screens applicants and coordinates interviews, plus an agentic AI layer (Salli) for scheduling and labor-law compliance.

These are company details, not findings, and carry no weight in the score.

## Head-to-head comparisons

- [Harri vs Recruitee](https://www.hireaiscore.com/compare/harri-vs-recruitee): 41 vs 42
- [Harri vs softgarden](https://www.hireaiscore.com/compare/harri-vs-softgarden): 41 vs 42
- [Harri vs Cezanne Recruitment](https://www.hireaiscore.com/compare/cezanne-recruitment-vs-harri): 41 vs 41
- [Harri vs Recruitly](https://www.hireaiscore.com/compare/harri-vs-recruitly): 41 vs 41
- [Harri vs Radancy](https://www.hireaiscore.com/compare/harri-vs-radancy): 41 vs 41
- [Harri vs Manatal](https://www.hireaiscore.com/compare/harri-vs-manatal): 41 vs 41

## 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.

Casework has no commercial relationship with this vendor.

Full policy: https://www.hireaiscore.com/about

## How to cite

HireAIScore, "Harri" profile, rubric v1.0, last reviewed 2026-08-02. https://www.hireaiscore.com/vendors/harri

Please link the profile URL rather than this file, quote the score with its rubric version and review date, and include the vendor's reply where one is published.
