Calaris

Policy

AI Ethics Policy

Established: May 18, 2026

We aim to build a society in which people are evaluated not by their background alone, but by their Caliber — the genuine ability to create value in any environment.

We provide an AI assessment service that makes this "Caliber" visible, based on our proprietary evaluation framework.

At the same time, AI services are generally recognized as carrying risks such as unjust bias and discrimination based on diverse backgrounds, privacy infringement, and hallucination.

In light of these risks, we hereby declare this policy: we will strive to ensure fairness, reliability, and safety in our services, and take responsibility for realizing the society we envision.

Article 1: Human-Centric Principle

We develop and provide AI services based on reasonable human judgment, and comply with all procedures established to ensure the fairness, reliability, safety, and other essential qualities of our services.

Article 2: Ensuring Fairness and Safety

In developing and providing our AI services, we periodically verify evaluation criteria and output tendencies to prevent unjust discrimination or bias based on diverse backgrounds such as race, gender, nationality, or age, and strive to make corrections as necessary.

We also check for output anomalies, such as the generation of unintended information or sudden shifts in score distributions. If we detect an anomaly, we promptly analyze the cause and work to review settings, prompts, and related configurations.

Article 3: Privacy Protection

We recognize that the proper handling of users' personal information is an important social responsibility. Personal information we collect — including video data, audio data, and evaluation results — is managed and handled in compliance with the Act on the Protection of Personal Information and other relevant laws, as well as our Privacy Policy and Terms of Service. We strive to prevent unauthorized use and data breaches.

Article 4: Ensuring Security

We endeavor to implement reasonable security measures in accordance with current technical standards to prevent unintended behavior, changes, or information leaks caused by attacks on, or unauthorized manipulation of, our AI systems.

Article 5: Transparency and Accountability

For outputs related to evaluations and important operational decisions, we strive to keep records for a set period, in enough detail that those outputs can be reproduced and verified as needed. We also establish a system for responding appropriately to customer inquiries about our compliance with this policy.

Article 6: Ensuring AI Literacy

We strive to make this policy, relevant laws, and guidelines known to our officers, employees, and all other personnel involved in delivering our services, and to take the measures needed to ensure a sufficient level of AI literacy.