Responsible Artificial Intelligence Policy

Purpose and Scope

Türkiye Sigorta adopts an ethical, secure, transparent, and sustainable approach to the development, procurement, and use of artificial intelligence technologies. This policy aims to ensure that artificial intelligence systems are managed in a manner that respects human rights, complies with regulations, and adheres to corporate responsibility principles throughout their entire lifecycle.

The Company has established an Artificial Intelligence Management System aligned with ISO 42001 to ensure the ethical, transparent, and trustworthy use of AI applications. Within this framework, the Company adopts a holistic approach to governance, risk management, data management, and model lifecycle processes. Compliance requirements under ISO 42001 are regularly reviewed and reported.

The Company also applies the same ethical, security, and compliance standards to the use of artificial intelligence systems developed, provided, or operated by third parties. Suppliers and business partners are evaluated for compliance with principles of data privacy, cybersecurity, fairness, transparency, and human oversight; risk-based third-party audits are conducted, and systems that do not meet these requirements are not used.

1. Fundamental Principles

1.1 Data Privacy and Protection

The Company places the utmost importance on the protection of personal data throughout the design, development, and operation of artificial intelligence systems. In this context, all processes are conducted in compliance with relevant data protection legislation (KVKK), and data privacy is safeguarded in accordance with the principles of data minimization and anonymization.

1.2 Cybersecurity

The cybersecurity of artificial intelligence systems is a fundamental component of corporate risk management. Systems are protected against unauthorized access, data breaches, and misuse through a high level of security controls. Accordingly, secure software development standards are applied, regular penetration tests are conducted, and incident response mechanisms are actively maintained.

1.3 Fairness and Mitigation of Bias

The Company ensures the identification and mitigation of biases that may arise in artificial intelligence models. The diversity of training datasets is ensured, fairness and discrimination risks are regularly assessed, and corrective actions are taken on model outputs when necessary.

1.4 Human Oversight

Human oversight is essential in critical decision-making processes. Artificial intelligence systems are not permitted to make irreversible or high-impact decisions autonomously. Humans are involved in the process with the authority to approve, monitor, or intervene when necessary, and systems are designed to always allow for human intervention.

1.5 Transparency and Explainability

The Company adopts transparency and explainability as fundamental principles in the use of artificial intelligence. When artificial intelligence systems are used, this is clearly communicated to relevant stakeholders, and system outputs are grounded in justifications that are as understandable as possible. Situations where users interact with artificial intelligence are clearly indicated.

1.6 Accountability

Responsibility for all outputs and decisions arising from AI systems is clearly defined. In this context, relevant business units and authorized managers are responsible for system performance, risk management, and compliance processes. Formal processes are established to investigate and address AI-related errors or adverse effects. The Artificial Intelligence Management System is subject to internal audit at least once a year.

1.7 Usage Boundaries

The scope of use for AI systems is clearly defined, and the use of systems beyond their designated purposes and authority limits is not permitted. The capacity, limits, and usage conditions of the systems are clearly defined to prevent risks of misuse and unauthorized use.

1.8 Environmental Impact

The Company takes into account the environmental impacts of artificial intelligence systems and prioritizes solutions with a low carbon footprint. Energy consumption, data center efficiency, and resource usage are regularly assessed; the use of technologies that enhance energy efficiency is encouraged, and the shift toward renewable energy sources is supported.

1.9 Prohibited and Unethical Uses

The Company does not develop or use the following artificial intelligence applications:

This scope is aligned with international best practices and relevant regulations, and has been determined with reference to prohibited AI applications defined under the European Union AI Act (EU AI Act).

RESPONSIBILITY

This Policy has been approved by the Board of Directors of Türkiye Sigorta A.Ş. and is periodically reviewed in light of changes in legislation, the Company’s activities, and evolving governance expectations.