On 11 May 2026, the Artificial Intelligence Centre approved the inclusion of PPPA's application in the special regulatory sandbox, established pursuant to Section 8 of the Law on the Artificial Intelligence Centre and Article 57 of the EU Artificial Intelligence Act.
What Process-as-Code Is
Process-as-Code is a platform developed by PPPA that converts an organisation's internal processes, decision logic and regulatory requirements into a machine-readable, version-controlled form with a full audit trail. On top of this data layer, artificial intelligence agents can prepare or make decisions in typical public administration processes — reviewing applications, classifying documents, issuing permits, prioritisation — while retaining human oversight of the final decision.
Why This Is Happening in a Regulatory Sandbox
Under the EU Artificial Intelligence Act, AI systems that support public-sector decision-making are classified as high-risk. They require documented risk management, technical documentation, human oversight and a conformity assessment. A regulatory sandbox is a controlled framework in which such a system can be tested with real data prior to wider deployment — in cooperation with the competent authorities.
What Happens Next
The planned duration of operation within the regulatory sandbox is 3 to 6 months. During this period, together with the Latvian AI Centre and the competent institutions, PPPA will:
- validate the quality of AI decisions using real public-sector data;
- explore the boundaries between automated and semi-automated decisions;
- map the AI chain of trust from the algorithm to the specific decision;
- document the results as a knowledge base for future AI and PPP projects.
The results will be published as an open compendium of lessons learned that Latvia's state administration, municipalities and private sector can use to structure their own AI projects.