Softtek Advocates Transparent AI in Justice and Public Sectors
Softtek has analyzed the key factors for more transparent and understandable artificial intelligence in the justice and public sectors. The company highlights the need for explainable AI (XAI) and trust-centric design to address the growing demand for algorithmic transparency.
The Push for Transparency in AI
The adoption of AI in public administration and justice has reached a turning point. Citizens and legal professionals are calling for an end to algorithmic “black boxes”. This shift is driven by the European AI Act, which mandates comprehensible, auditable, and justifiable systems.
Softtek’s White Paper, released for the DATAfórum Justicia 2026, emphasizes the importance of XAI as a technical foundation for auditing algorithmic behavior. The company also introduces the concept of Trust-Centric Experience, which focuses on translating complex technical processes into clear interfaces that convey control, security, and legal integrity to users.
Key Aspects of Transparent AI
To achieve transparency, Softtek outlines several critical aspects. First, moving beyond the “black box” model requires traceability. This involves combining XAI’s technical capabilities with trust-centric design to ensure visibility and clear justification for each algorithmic decision.
Compliance with regulations is another key factor. Integrating transparency, explainability, and disclosure principles from the design phase reduces friction with the AI Act and audit processes. This approach protects organizations from severe penalties and safeguards their return on investment.
Balancing Technical Complexity and User Experience
While XAI provides technical support for auditing, the Trust-Centric Experience focuses on the user interface. It aims to present algorithmic logic in a clear and understandable manner, giving users control, security, and the ability to intervene when necessary.
Dynamic governance and real-time control are also essential. Trust-centric design incorporates feedback mechanisms, uncertainty management, and data traceability, replacing static policies with adaptive, risk-based governance.
Doris Seedorf, CEO of Softtek Spain, notes that users demand immediate results but reject anything they cannot understand.
By prioritizing transparency and explainability from the design stage, organizations can overcome adoption barriers in regulated public sector environments. This approach standardizes trust capabilities, enabling the rapid, secure, and consistent scaling of AI solutions.
