More and more organisations are adopting AI. But governance, accountability and risk management are often left unresolved. We balance growth with risk management.
The EU AI Act and sector regulation (DORA, NIS2, GDPR) create obligations that cannot be delegated to the legal department alone. Accountability always stays at management level.
AI failures are not theoretical. Reputational, legal and operational damage is tangible, and organisations often recognise it too late.
Systems that act autonomously raise questions of accountability that no existing framework yet answers. That has to be prepared for now.
Vija Kalniņa is an expert in AI ethics, law and policy with more than ten years of legal experience in the private and public sectors. Her interdisciplinary approach combines legal precision with a deep understanding of technological reality.
The EU AI Act places an obligation on organisations: every employee who comes into contact with AI must have appropriate training. We make sure your team not only meets that requirement but is also able to work with AI meaningfully.
AI governance for decision-makers, making it possible to lead AI adoption within a sound legal framework while preventing reputational, legal and financial risks.
Maturity assessment, mapping the gaps in AI governance, developing guidelines, forecasting risks, building an ethical framework.
Commissioned by VARAM and produced in collaboration with MILA, our experts developed "Practical guidelines for AI adoption in the work and services of public institutions" – the first state-commissioned framework of its kind in Latvia.
The document translates the language of regulatory requirements into concrete, usable steps for public institutions – from risk assessment through to protocols for responsible AI use.
⚑ Placeholder – quote from a public sector representative on the practical value of the guidelines for public institutions.
If you're a state or municipal institution, the "Practical Guidelines for AI Implementation in Institutional Work and Services" VARAM published in late 2025 are a real, concrete starting point. But the guidelines themselves are only a framework: the decisive step is adapting them to your institution's situation and sector specifics, which is where our experts can help.
There's no single, sector-wide set of guidelines for the private sector — the baseline is the EU AI Act, most of whose requirements took effect on 2 August 2026. We help translate those regulatory requirements into an internal AI policy for your organisation that fits both the Act and your industry's specifics.
Yes. Our team includes EU law specialists who work directly on compliance, data protection and GDPR — they'll help you understand how your AI use measures up against the requirements, and prepare the internal policy your organisation needs.
That's very common — Gartner's 2026 survey found that 88% of employees with access to company AI tools also use personal AI tools for work tasks, which creates a real risk of sensitive information leaking out. We start with an organisational audit to understand how AI is already being used, and build a realistic, adoptable framework from that — not a theoretical document nobody reads.
No. We build the framework alongside your ongoing work — the process covers research, drafting, legal validation, rollout with training, and then regular maintenance, but your organisation keeps using AI throughout, with an increasingly clear sense of direction.
It depends on your organisation's size and complexity — we start with an audit and risk assessment, followed by drafting policy and standards, validating with employee groups, and rollout. We set a concrete timeline after the first conversation.
Yes. The EU AI Act requires that every employee who works with AI receive appropriate training — that's built into our AI training programmes, which we tailor for both leadership and specialists.
Our methodology is built on real work with public-sector organisations, academic research and current international expertise — we're the team that developed Latvia's first state-commissioned AI guidelines framework. You can build this competence in-house too, but it takes time organisations often don't have — we already know the regulation and its practical application.
Book a free 30-minute consultation. We'll get to know your organisation's situation and recommend the best first step.