EOS leverages AI – safely and responsibly.

In debt collection, EOS leverages AI to classify incoming emails, swiftly identify senders’ concerns, and route each case accurately. EOS also leverages call transcripts and analytics to support call centers and collections agents. This is made possible by the generative AI assistant SofiK on the K+ platform, a receivables management platform developed by EOS in Romania. The secured and unsecured investments teams also use conventional machine-learning models to produce data-driven forecasts and valuations that inform pricing for NPL portfolio acquisitions, for example.

AI already enhances many internal functions today – from proactively defending against cyberattacks and boosting productivity and information processing to supporting the legal department with AI-assisted contract drafting and review. In the area of Group Governance, we deploy AI selectively to create concise, up-to-date overviews of the cyber threat landscape and to streamline policy management.

Furthermore, we harness AI to craft communications content and handle translations. We are planning to expand into regulatory analysis, risk and audit processes, basic Know Your Customer (KYC) checks, and data discovery to enhance data protection.

As a first step, the EOS AI policy establishes a binding framework aligned with the regulatory requirements of the EU AI Act and the GDPR. In accordance with policy, all planned AI applications must be reviewed before use. This review is carried out by the relevant subject-matter owners and/or AI users, in coordination with local data protection officers and management. In addition, Group Governance also oversees Group-wide procedures.

The data protection teams in the country organizations and/or at EOS Holding then perform a risk-based review of each AI application, classify it, and, where necessary, assign it to the regulated categories under the AI Act. The policy also defines internal approval processes and delivers actionable guidance. Clearly defined roles, responsibilities, and processes ensure the safe use of AI.

Empowering employees is another key pillar of our AI compliance framework. Governance, compliance, cybersecurity, and data‑protection experts at the Holding and country levels act as multipliers, proactively sharing their expertise throughout the organization. We also educate every employee with targeted information on the responsible use of approved AI tools.

The Group Governance department sets the framework for the use of AI and ensures uniform standards across all countries and business units. Local managing directors and data protection officers in each country organization are responsible for implementing EOS AI policies on-site, particularly for evaluating and documenting specific AI use cases. Through our governance community, we stay closely connected with our specialist colleagues, enabling us to support and monitor ongoing developments. This ensures that we rigorously monitor each of our business models and the AI risks they entail, from our established fiduciary services to our activities as an NPL investor.

We see AI governance as a dynamic process. This means we continuously monitor both technological advances and evolving regulatory requirements. By working closely with internal specialists and consulting external experts and authorities, we maintain a flexible framework that enables us to respond proactively to new challenges. However, our goal is not only to respond but to identify trends at an early stage and drive innovation. We currently see AI’s greatest potential in coding, underwriting, and optimizing our support functions.