Responsible AI & the EU AI Act

A human decides — and we make reviewing everyone genuinely easy

Mokka sources, screens, and evaluates candidates, then hands the recruiter a clear, evidence-backed view of every one of them — so a person makes each call. Mokka never makes a hiring decision on its own.

Every vendor says "a human is in the loop." That only means something if reviewing every candidate is genuinely easy — so we built the product around it:

• Candidates are organized into clear, reviewable categories — nothing sits in an unread pile.
• Split-view and side-by-side layouts let recruiters compare candidates and their evidence at a glance.
• Every assessment is broken down requirement by requirement, with the supporting evidence quoted, plus scores and flags that surface exactly what needs a closer look.

The result: recruiters review the whole pipeline with confidence, instead of rubber-stamping whoever floated to the top.

Scored on evidence, not identity

• We don't collect or infer a candidate's protected characteristics (race, sex, ethnicity, age, disability), and we don't use proxies such as postal code, address, or other location data in scoring. Candidates are assessed on the evidence they provide against the recruiter's own role criteria.
• Candidates choose how to respond — text, voice, or video — with no time pressure, can opt out of the AI assessment and still be considered, and can request human review at any point.
• We score the content of what a candidate says — never voice, accent, or appearance — and we perform no voiceprint matching or facial recognition.

Independently audited for fairness

We commission an independent third-party bias audit of our candidate scoring. It measures selection rates and impact ratios across sex, race/ethnicity, and intersectional groups, and Mokka shows no statistically significant adverse impact. The audit is built to satisfy the EU AI Act's fairness and non-discrimination expectations, and it also covers the four-fifths adverse-impact standard used under US frameworks such as New York City's Local Law 144. Customers and their counsel can review it through our Trust Center.

Privacy & data

Mokka is SOC 2 Type II and Cyber Essentials certified, GDPR and CCPA compliant, and independently penetration-tested. We do not use customer or candidate data to train our models. Scoring runs on established foundation models guided by explicit criteria — so no one's historical hiring patterns are ever baked into a model.

The EU AI Act

AI used to evaluate candidates is classified as high-risk under the EU AI Act, and Mokka meets that standard:

• Risk management & data governance maintained for the scoring system, with the independent bias audit as standing fairness evidence.
• Technical documentation kept current for the system, its design, and its performance.
• Human oversight by design — a recruiter decides, nothing is auto-rejected, and every decision is logged.
• Transparency — explainable, evidence-cited assessments a recruiter can verify against the source.
• Accuracy & robustness — deterministic scoring, version control, and continuous monitoring.

Conformity for recruitment AI is self-assessed by the provider, and Mokka maintains the documentation that demonstrates it. For customers who need to complete a Fundamental Rights Impact Assessment, our audit and documentation provide the inputs.

What you can get

• Public: this statement, and summaries of our fairness audit and EU AI Act posture.
• On request, under NDA via our Trust Center: the SOC 2 Type II report, the full independent bias audit, and the detailed conformity documentation for your security and legal teams.

Trust Center: trust.gomokka.com · Questions: privacy@gomokka.com

Mokka is a product of Mokka Labs Ltd. This statement describes our current posture and is updated as needed.