We help businesses and institutions use artificial intelligence in a way that withstands regulatory scrutiny — from mapping the AI tools they use to preparing the documentation required under the AI Act.
The European Union's Artificial Intelligence Act (AI Act) applies directly in all Member States, with successive sets of obligations taking effect in stages. In Poland, it is supplemented by national legislation on artificial intelligence systems, which establishes domestic supervision and enforcement rules.
Determining Roles and Classifying AI Systems
The scope of obligations depends on two factors: the role an organisation plays and the system's risk category. Different requirements apply to AI system providers and to organisations that deploy AI systems in their own operations, and the distinction between these roles can be subtle — a substantial modification of a tool or making it available under one's own brand may be enough to change an organisation's role. We conduct an inventory of AI solutions in use, assign risk categories and roles, and document the results in a register that serves as the starting point for all subsequent compliance obligations.
Prohibited AI Practices and Transparency Obligations
Certain uses of artificial intelligence are prohibited outright — including social scoring, manipulation that exploits individuals' vulnerabilities, and specific forms of emotion recognition and biometric categorisation. Transparency obligations form a separate category: informing individuals that they are interacting with an AI system and labelling AI-generated content. These obligations also apply to organisations that use only publicly available AI tools — which is why we address them first, before moving on to more advanced requirements.
High-Risk AI Systems
AI solutions used in recruitment and employee management, creditworthiness assessments, education, access to public services and healthcare are subject to the most extensive requirements: risk management systems, data quality standards, technical documentation, record-keeping, human oversight and post-market monitoring. The application dates of some of these obligations have been postponed under an EU package of legislative amendments, giving organisations additional time to prepare without removing the requirements themselves. We help organisations make effective use of this time by developing implementation plans, preparing documentation and allocating responsibilities between AI system providers and deployers.
AI Usage Policies and AI Literacy Obligations
The obligation to ensure an appropriate level of AI literacy among people using AI systems applies to every organisation, regardless of the system's risk category. We prepare artificial intelligence usage policies covering permitted uses of generative AI tools, types of data that must not be entered into them, methods for verifying outputs and procedures for documenting the use of AI in legally significant processes. We provide training for management boards, HR departments, marketing teams and operational staff, tailored to the tools they actually use.
Contracts and Procurement of AI Solutions
We review and negotiate contracts for the supply and implementation of artificial intelligence systems, addressing the allocation of compliance responsibilities, warranties concerning training data, rights to system-generated outputs, liability for incorrect decisions, supplier audits and exit arrangements. We advise public sector entities on preparing procurement procedures for AI-based solutions — from drafting descriptions of the subject matter of contracts to establishing tender evaluation criteria and contractual clauses reflecting AI Act requirements.
The Intersection of AI, Data Protection, Employment Law and Intellectual Property
Implementing AI is rarely a matter of AI Act compliance alone. Recruitment systems raise employment law and personal data protection issues, generative AI tools involve copyright in input materials and generated outputs, and diagnostic solutions are subject to medical device regulations. We address these interconnected issues within a single team rather than dividing them into separate legal opinions that are difficult to translate into a coherent business decision.
Artificial Intelligence Law – AI Act, Implementation and Compliance - Styczynski-Keller
