AI Policy

Last updated: December 2025

ARTIFICIAL INTELLIGENCE MANAGEMENT SYSTEM (AIMS) POLICY

1. Our Commitment

We are committed to developing and using Artificial Intelligence in a responsible, safe, and legal manner. An Artificial Intelligence Management System (AIMS) has been established in accordance with ISO/IEC 42001 to manage risks, protect data, and ensure quality, transparency, and user trust.

2. Key principles

In all our activities connected to AI, we apply the following principles:

Compliance with applicable regulations, contracts and internal rules.
Managing the risks and impacts of AI systems through assessments, controls, and approvals.
Data protection and privacy - data minimization, access control, security measures.
Security and reliability of the solution - testing, verification, operational monitoring, change management.
Transparency and traceability - clear purpose, limitations, record of decisions and changes.
Fairness and quality - control of the quality of data and results, mitigation of bias.
Human supervision when necessary, proportionate to the risk and purpose of the system.

3. Governance and responsibilities

The company's management approves this policy and provides resources for AIMS. The roles and responsibilities for the management, development, security, compliance, and operational monitoring of AI systems are clearly defined and regularly reviewed.

4. Third-Party Vendors and AI Tools

We use AI services, clouds and third-party tools only with prior assessment and approval, with contractual and security requirements - data protection, processing conditions, data retention limits, to the extent applicable.

5. Incidents and continuous improvement

We have implemented a process for reporting, managing incidents, non-compliances, corrective actions, and work continuously to improve AIMS through objectives, performance assessment, internal audits and a periodic management review.

6. Contact

For questions about this policy and our approach to AI management, please contact us via our contact form.

Reference: AIM-SOP-06-FRM01 Release: 1.0 - Released: 15.12.2025

POLITIKA SISTEMA UPRAVLJANJA UMJETNOM INTELIGENCIJOM (AIMS)

1. Naša posvećenost

Posvećeni smo razvoju i korištenju umjetne inteligencije na odgovoran, siguran i zakonit način. Sistem upravljanja umjetnom inteligencijom (AIMS) uspostavljen je u skladu sa ISO/IEC 42001 kako bismo upravljali rizicima, zaštitili podatke i osigurali kvalitet, transparentnost i povjerenje korisnika.

2. Ključni principi

U svim AI aktivnostima primjenjujemo sljedeće principe:

Usklađenost sa važećim propisima, ugovorima i internim pravilima.
Upravljanje rizicima i uticajima AI sistema kroz procjene, kontrole i odobrenja.
Zaštitu podataka i privatnosti - minimizacija podataka, kontrola pristupa, sigurnosne mjere.
Sigurnost i pouzdanost rješenja - testiranje, verifikacija, nadzor u radu, upravljanje promjenama.
Transparentnost i sljedivost - jasna namjena, ograničenja, evidencija odluka i promjena.
Pravednost i kvalitet - kontrola kvaliteta podataka i rezultata, ublažavanje pristrasnosti.
Ljudski nadzor kada je potrebno, proporcionalno riziku i namjeni sistema.

3. Upravljanje i odgovornosti

Management kompanije odobrava ovu politiku i osigurava resurse za AIMS. Uloge i odgovornosti za upravljanje, razvoj, sigurnost, usklađenost i operativno praćenje AI sistema su jasno definisane i redovno preispitivane.

4. Dobavljači i AI alati trećih strana

AI servise, cloud i alate trećih strana koristimo samo uz prethodnu procjenu i odobrenje, uz ugovorne i sigurnosne zahtjeve - zaštita podataka, uslovi obrade, ograničenja zadržavanja podataka, u mjeri koja je primjenjiva.

5. Incidenti i kontinuirano poboljšavanje

Imamo uspostavljen proces za prijavu i upravljanje incidentima i neusklađenostima, provodimo korektivne mjere. AIMS kontinuirano unapređujemo kroz ciljeve, mjerenje učinka, interne provjere i periodični pregled rukovodstva.

6. Kontakt

Za pitanja o ovoj politici i našem pristupu upravljanju AI, kontakt:

Contact form

Oznaka: AIM-SOP-06-FRM01 - Izdanje: 1.0 - Datum izdanja: 15.12.2025