Arctic Shores
Arctic Shores sells a task-based psychometric assessment platform ("Una") that scores candidates on behaviour captured during short interactive tasks — covering personality, workplace intelligence and learning agility — rather than self-report questionnaires or CV review.
§ 01 — Score breakdown
§ Score breakdown
Category scoring
Weighted contribution shown to the right of each bar.
- 01
Article 11 Technical Documentation
Weight 20%50
+10.0
- 02
Bias Audit Transparency
Weight 18%35
+6.3
- 03
FRIA Support
Weight 15%35
+5.3
- 04
Data Governance Disclosure
Weight 15%57
+8.5
- 05
Human Oversight Design
Weight 12%56
+6.7
- 06
Post-Market Monitoring
Weight 12%30
+3.6
- 07
Customer Documentation
Weight 8%64
+5.1
§ 02 — Strongest · weakest
Strongest category
Customer Documentation
Raw score 64 · contributes 5.1 to total.
Weakest category
Post-Market Monitoring
Raw score 30 · contributes 3.6 to total.
§ 03 — Cited evidence
Download diligence record→§ Evidence
Cited per category
Every score is backed by at least one cited piece of evidence.
§ 04 — Editorial notes
Company overview
Arctic Shores Limited (Companies House no. 08589048) was incorporated on 28 June 2013 and is registered at Lowry House, 17 Marble Street, Manchester. Its platform, branded "Una", replaces question-based psychometrics with interactive behavioural tasks: candidates are scored on what they do during the task rather than on what they say about themselves, producing a profile of what the company markets as "Skill-enablers" spanning personality, workplace intelligence and learning agility. The company reports use by over 350 organisations and publishes integrations with Greenhouse, SmartRecruiters, Teamtailor, Tribepad, Willo and Eploy alongside a standard API. Two positioning claims recur across the site: that the assessment is resistant to manipulation by generative AI, and that it is "never a black box". Note for the record: no product, module or feature named "Silhouette" could be located anywhere on arcticshores.com or in public sources, and that name should not be attributed to this vendor without further confirmation.
Regulatory exposure
As a provider of an AI-assisted candidate evaluation tool, Arctic Shores sits squarely in Annex III high-risk territory under the EU AI Act, with obligations landing from 2 August 2026, and it sells actively into European markets. Its strongest defensible position is architectural rather than documentary: the company consistently frames the product as decision-support that "never decides for you", which materially reduces (without eliminating) UK/EU GDPR Article 22 automated-decision exposure and supports a meaningful-human-oversight argument. Its weakest position is evidentiary. The company asserts on its science page that its assessment produces "no adverse impact, regardless of ethnicity or gender", and asserts on a product page that a task is "GDPR and EU AI Act compliant" — yet publishes no adverse-impact data, no independent bias audit, no technical documentation pack, no conformity-assessment reference, and no NYC Local Law 144 audit summary. Arctic Shores holds ISO 27001 but not ISO 42001, and its published compliance and security material is information-security-shaped, not AI-governance-shaped. US deployers in New York City, Illinois and Colorado would need to generate the required artefacts themselves, since the vendor supplies none of them publicly.
Path to a higher score
The single highest-value move is to publish an adverse-impact analysis — ideally an independent audit by a recognised auditor — carrying sample sizes, selection rates by race/ethnicity and sex, four-fifths-rule impact ratios and a dated scope statement. The unsupported "no adverse impact" claim is currently the widest gap between what Arctic Shores says and what it evidences, and closing it would move two rubric categories at once. Second, publish a technical documentation pack: a system/model card describing which behavioural signals are captured, how they are combined into scores, norm-group composition, validity and reliability coefficients and known limitations, plus Article 13-style instructions-for-use so deployers can meet their own obligations. Third, add explicit EU AI Act deployer guidance including an Article 27 FRIA template and a per-jurisdiction note covering NYC LL 144, Illinois and Colorado. Fourth — the cheapest fix and the lowest-scoring dimension — stand up a public post-market surface: a status page, a product/model changelog and a published vulnerability disclosure policy with a security contact. ISO 42001 certification would consolidate the first two and lift the technical-documentation and data-governance scores together.
§ Compare
Build any comparison→Arctic Shores against its nearest-scoring peers.
In Assessment AI.
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.