Why the public sector cannot act alone
AI implementation requires three sets of resources: technology, data and competence. The public sector often has the data but lacks the technology and specialised competence. Classic procurement does not solve this problem — it acquires a product but does not build capacity. The PPP model is structurally different: the private partner takes on responsibility not only for development, but also for operation, maintenance and outcomes.
Estonia's example: from stenographers to AI
In Estonia, a PPP between the technology company Finestmedia and Tallinn University of Technology created the voice recognition system "Hans", which produces verbatim transcripts of parliamentary sittings. The system was trained on a base of 1,500 hours of recordings and integrated into the parliament's information systems. The result: the complete replacement of stenography services with a more accurate and scalable solution. This is a classic case of PPP in digitalisation — the private partner's technological competence combined with state infrastructure and data.
In parallel, Estonia is implementing the national AI action plan "Kratt" for 2024–2026, which directly provides for public-private cooperation mechanisms for implementing AI in public services, including the adaptation of procurement procedures and certification schemes.
Latvia: the next step is achievable
Latvia currently has no structured policy for implementing AI in the public sector through PPP mechanisms. This creates a risk: the technological gap with neighbouring countries is widening. But it also creates an opportunity: the first large-scale project can be defined now — in document management, procurement analysis, or the automation of municipal services.
PPPA invites state institutions, ministries and municipalities to begin a structured dialogue with the private sector on opportunities for implementing AI in a PPP format.