Contact

Safe AI in your business. Move faster, while staying in control.

AI is already being used in your organisation-in employees’ tools, suppliers’ systems, HR, marketing, sales and customer-service processes, and in products. The question is: do you have it under control?

Book a call about AI compliance

AI is developing faster than the rules intended to govern it

In many companies, AI is no longer a plan for the future. It is part of day-to-day operations. Employees use ChatGPT, Copilot, content generators, HR tools, sales automation, scoring, analytics and suppliers’ systems.

The problem begins when no one has a complete picture: where AI is being used, what data it processes, whether it affects decisions, who is responsible for the risk, and which uses may trigger regulatory obligations.

“The greatest AI risk in a company does not always arise from a major implementation. It often starts with a tool that no one has formally approved but that is already affecting data, customers, employees or business decisions.”

AI Compliance is not intended to halt innovation. It gives the organisation clear rules so that the management board, managers, IT, HR, legal, compliance and operations teams can make decisions faster-without recurring confusion or shifting responsibility.

Mateusz Sawaryn

Managing Partner Counsel

Common challenges

The most common problems we solve

You do not know where AI is actually being used in your business Employees test tools, departments buy solutions from suppliers, and some AI features appear in systems the organisation already uses. The management board sees the trend but not the organisation’s full risk exposure.

We create a map of AI use across the organisation: tools, processes, departments, suppliers, data, decisions and accountable owners.

You gain visibility of what is already happening and can prioritise action where the risk is real.

AI will otherwise develop from the bottom up, without control over data, confidentiality, IP, accountability or its impact on customers and employees.

You do not know which uses are subject to the AI Act Not every use of AI entails the same level of obligations. A marketing content generator differs from an HR tool, a scoring system, an AI-based SaaS product or a solution that affects decisions concerning customers.

We classify AI uses by reference to the AI Act, the organisation’s role, the level of risk and the obligations that may apply to the company.

You receive a practical decision framework: permitted; permitted subject to conditions; approval required; high risk; not recommended.

The company may implement solutions whose risks have not been assessed-or block safe projects simply because clear criteria are lacking.

Each department uses AI in its own way Marketing generates content. HR tests recruitment tools. Sales automates customer contact. IT implements integrations. Compliance and Legal learn about everything after the event.

We design an AI policy, rules for using tools, a model for approving new implementations, checklists and rules for documenting decisions.

The organisation operates under a single model. Teams know what they may do independently, when approval is required and whom to involve.

Otherwise, operational disorder follows: duplicated questions, unclear decisions, delays, conflicts between departments and risks with no owner.

Employees want to use AI but do not know what is prohibited The greatest risks often arise in simple situations: entering customer data into an AI tool, using confidential information, generating content that creates IP risk, or basing a decision on an output that no one has verified.

We prepare practical rules for employees and managers, together with an AI literacy programme tailored to roles within the company.

People can use AI with greater confidence because they understand the boundaries: data, confidentiality, IP, verification of outputs, accountability and escalation.

Otherwise, the organisation leaves AI security to employees’ day-to-day decisions-without a common standard or predictability.

Customers, investors or auditors are asking about AI governance Enterprise customers increasingly ask suppliers, partners and portfolio companies about their rules for using AI. The absence of a policy, register, approval process and documentation can undermine credibility in sales, due diligence or an audit.

We help prepare documents, processes and responses demonstrating that the company uses AI consciously, responsibly and systematically.

AI governance becomes a strength in discussions with customers, investors and partners-not a problem to explain at the last minute.

The company may lose momentum in a sales, investment or audit process because it cannot demonstrate how it manages AI.

The management board is responsible for risks it cannot see Without an AI map, clear rules and a decision-making model, the management board has no control over where AI affects the company’s data, processes, customers, employees and reputation.

We establish a governance framework: roles, responsibilities, approval thresholds, a register of AI uses, rules for documenting decisions and monitoring mechanisms.

The management board gains control without having to approve every case manually. Decisions are visible, documented and embedded in the process.

Otherwise, the organisation makes technology decisions without proper oversight-and responsibility returns to management only when an incident occurs.

Our solution

An AI Compliance Operating System for your organisation

We do not create documents that sit on a shelf. We build a practical system for using AI: from a map of use cases and risk assessment to rules for people and a process for approving new tools.

Map of AI use across the organisation

  • identification of current and planned uses of AI,
  • review of internal tools, suppliers’ systems and automation,
  • analysis of HR, sales, marketing, customer service, IT, product and analytics processes,
  • identification of process owners and decision points.

 

  • AI use map
  • list of priorities

AI Act classification and organisational roles

  • assessment of whether particular solutions may fall within the scope of the AI Act,
  • determination of the organisation’s role: provider, deployer, user, importer, distributor or another role depending on its operating model,
  • preliminary risk-level classification,
  • identification of obligations and areas requiring particular oversight.

 

  • classification of AI uses
  • list of obligations assigned to specific processes

Mapping legal and business risks

  • the interaction between the AI Act and GDPR, cybersecurity, IP, confidentiality and trade secrets,
  • assessment of the impact of AI on decisions concerning customers, employees or candidates,
  • analysis of contractual, reputational and management risks,
  • identification of critical and acceptable risks, and risks that can be mitigated through straightforward rules.

 

  • risk matrix

AI policy and rules for using tools

  • preparation of a practical AI policy,
  • rules for working with personal data, confidential information, content and know-how,
  • rules for verifying AI outputs,
  • rules for using public and company tools,
  • communications for employees and managers.

 

  • an AI policy and a set of rules that the business can understand

Approval model for new AI implementations

  • a process for submitting new AI tools and uses,
  • risk-assessment criteria,
  • decision thresholds and approval paths,
  • the roles of Legal, the DPO, IT, Security, HR, Compliance and the business,
  • checklists for managers and project teams.

An AI approval procedure that speeds up decisions and reduces ad hoc consultations.

Documenting decisions and oversight

  • a register of AI uses,
  • tool-assessment forms,
  • decision-documentation templates,
  • monitoring and review rules,
  • mechanisms for updating rules following technological or regulatory changes.

A package of documents and tools for the organisation’s ongoing oversight of AI.

AI literacy and implementing the rules across the team

  • training for employees and managers,
  • practical AI-use scenarios,
  • rules for escalation and raising questions,
  • implementation materials for teams,
  • support in communicating the rules across the organisation.

The team knows how to use AI safely, when to seek approval and how to avoid the most common mistakes.

Who this service is for

For companies that want to develop AI without decision-making disorder

This service is for organisations that already use AI or are planning a broader implementation and need clear rules, accountability and a practical operating process.

Head of Legal / Compliance Officer

when the AI Act, GDPR, cybersecurity, IP and business liability need to be translated into rules the organisation will actually use.

CTO / Head of Product / IT Director

when the company is implementing AI tools, building an AI-based product or needs a process that will not hinder technological development.

CEO / management board member / COO

when control is needed over how AI enters the business, who makes decisions and which risks require oversight.

Stages of our work

How we work together

Map AI use

We identify where AI is already used, who uses it, which data or decisions are involved and which areas need priority.

Classify obligations and risks

We assess AI use under the AI Act, GDPR, cybersecurity, IP, confidentiality, contractual liability and reputational risk.

Design policies and operating rules

We prepare an AI policy, procedures, checklists, an approval model for new deployments, decision-recording rules and core oversight mechanisms.

Implement, train and support

We help communicate the rules, train staff and managers, answer questions and establish ongoing monitoring.

Schedule a call about AI compliance.

Contact us

Bring AI under control before an incident, audit or enterprise customer does it for you

AI can accelerate business growth, but only if the organisation knows where it uses AI, what obligations apply, who makes decisions and how risks are documented.

Complete the short form. We will ask about current and planned uses of AI, your sector, the size of your organisation and the principal issue: regulation, data, product, HR, suppliers, enterprise customers or the absence of internal rules.

Mateusz Sawaryn

They will follow up with any questions or recommend a next step.

Quick response

We usually respond within two hours on business days.

    By submitting this form, you acknowledge that the personal data you provide will be processed by Kancelaria Prawna Sawaryn i Partnerzy sp.k. to handle your enquiry and contact you about it. Details are available in our Privacy Policy. The rules governing the use of this website are set out in the Terms and Conditions.