AI Governance Overview

This document summarizes how Kempian governs its use of artificial intelligence. It is for candidates, customers, and anyone evaluating how Kempian builds and oversees the AI Features.

*This document is part of Kempian's Trust Center documentation. It is reviewed periodically and does not constitute legal advice. Draft v0.2 — pending final legal review before publication.*

Kempian ("we," "us," "our") refers to Adept AI Inc., a Delaware corporation, principal place of business at registered address placeholder.

Our Approach to AI Governance

Kempian is an AI-assisted recruitment marketplace connecting employers, recruiters, and candidates. AI is built into how the Platform works — helping match candidates to roles, extract information from resumes, draft job descriptions, and answer questions inside the product.

Our governing philosophy is simple: AI should support human decision-making in hiring, not replace it. Recruiters and employers make the decisions that affect a candidate's livelihood — not algorithms. AI's role is to surface relevant information faster and more consistently, with an explanation attached, so the person making the decision is better informed.

Three principles run through Kempian's AI governance:

• Human oversight. AI outputs that could affect a candidate are designed to pass through a human review point before they take effect.
• Transparency. Candidates and customers are told when and how AI is used, and AI-generated outputs are labeled and explained rather than presented as unexplained conclusions.
• Accountability. Kempian, not its AI providers, is responsible for how AI is used on the Platform, and AI governance is owned at an organisational level within Kempian.

The AI Features

Kempian uses AI in four areas, collectively the AI Features:

• Candidate Matching — scores and ranks candidates against job requirements, with a "Why this match?" explanation of contributing factors so recruiters can evaluate the recommendation rather than take it on faith.
• Resume Parsing — extracts structured information (roles, dates, skills, education) from a candidate's CV, so profiles don't have to be built by hand.
• Job Creation Assistant — helps employers draft job description text from the details they provide.
• Chat Assistant — an in-platform conversational assistant that helps users navigate and use Kempian.

AI-generated outputs — match scores, rankings, and resume-derived profile fields — are recommendations only. Candidates and recruiters can always review and correct AI-derived data.

The Human Review Gate

Every AI output capable of affecting a candidate — a match ranking, an outreach message, a resume-derived profile field, or a generated job description — is designed to pass through a human recruiter or employer review point before it takes effect. This is the Human Review Gate. Recruiters can review, edit, and override AI outputs. Kempian's AI does not make final hiring, rejection, or shortlisting decisions.

Kempian is implementing human-confirmation checkpoints across all AI-assisted profile updates as part of its ongoing compliance program, so the Human Review Gate becomes a consistently enforced standard across every AI Feature. A material change to a production Candidate Matching model version triggers internal compliance review before deployment, and enterprise customers may request notification of material model changes as part of their agreement.

Our Regulatory Posture

Kempian's candidate-matching and resume-screening functionality is treated as high-risk AI used in employment under the EU AI Act. Kempian discloses AI use to affected individuals in line with the Act's transparency obligations, which are already in effect. This disclosure is provided through the AI Transparency Notice. Kempian is continuing to build out its broader AI Act conformity program ahead of the framework's later compliance deadlines.

Beyond the EU, Kempian's AI Features operate across a developing regulatory landscape in the United States and India, including state and sector-specific AI employment rules and India's Digital Personal Data Protection Act. Kempian monitors these frameworks and adapts its practices as they take effect, consistent with the jurisdiction-specific detail in the Privacy Policy and the AI Transparency Notice.

Kempian continuously enhances its AI governance and technical controls as part of its ongoing compliance program.

Standards Alignment

Kempian's AI governance and information security programs are designed with reference to the ISO/IEC 42001 (AI management systems) and ISO/IEC 27001 (information security management) frameworks. Kempian has not completed independent certification against either standard. Alignment with their control structures is part of Kempian's ongoing compliance roadmap.

Fairness and Explainability

Kempian does not claim its AI is bias-free — no vendor can honestly make that claim. Kempian is building a program to test for disparate outcomes in AI-assisted matching and screening. It also designs match explanations ("Why this match?") to show the confidence and contributing factors behind a recommendation, so a human reviewer can evaluate it rather than accept it automatically.

Want More Detail?

This overview is written for a general audience. Enterprise customers, prospective partners, and procurement or security teams conducting vendor due diligence can request Kempian's full AI Governance Statement under a mutual non-disclosure agreement, by contacting trust@kempian.com. The full Statement covers governance accountability, risk management, model governance, auditability controls, and a detailed regulatory-readiness roadmap.

Related Documents

This overview should be read together with: the AI Transparency Notice, the AI Acceptable Use Policy, the Privacy Policy, the Candidate Privacy & Visibility Notice, the Subprocessor List, and, for enterprise customers under NDA, the full AI Governance Statement.

Contact

• Enterprise due-diligence and procurement inquiries: trust@kempian.com
• General privacy questions: privacy@kempian.com
• Security vulnerability reports: security@kempian.com

Kempian — AI Governance Overview — v0.3 (Draft) — July 2026