At JAAQOB Holding, artificial intelligence tools support the work of our team, and in some of our services they form a deliberately designed part of the solutions we build for clients. In both cases the same principle applies: AI plays a supporting role and does not replace human knowledge, experience or responsibility. We work in line with our “Business First, Tech Second” philosophy – technology, including AI, is always a means of solving a specific business problem, never an end in itself.
The purpose of this notice is to explain the principles on which JAAQOB Holding uses AI tools in its internal work and in designing and delivering its services.
1. How we use AI
JAAQOB Holding uses AI tools with human control over the final outcome (human-in-the-loop). We rely primarily on general-purpose tools from established providers, and we assess planned uses against the requirements of Regulation (EU) 2024/1689 of the European Parliament and of the Council (the Artificial Intelligence Act, hereinafter: the AI Act) – in particular prohibited practices, uses classified as high-risk, and transparency obligations.
At JAAQOB Holding, AI does not make legal, financial, personnel or business decisions on its own. Decisions that affect people or clients are always made by a human.
2. Scope of AI support in our internal work
AI tools may support us in particular in the following areas:
- Analytical and editorial support – organising information, summarising materials, and language and structural support in drafting documents.
- Informational and marketing content – help in preparing materials on automation, technology and business strategy, published on our websites and social media.
- Visual layer – preparing graphic concepts and supporting visual elements.
- Meeting notes and summaries – transcription and summarisation tools support the documentation of internal conversations and conversations with clients. The rules for using these tools, including informing meeting participants, are set out in our internal procedures.
- Automation of operational processes – AI supports the organisation and classification of information in our internal processes (including correspondence handling, sales data and settlements), always in a supporting role, with decisions made by a human.
3. AI in services for clients
Designing and implementing automations with AI components is part of our offering. In projects of this kind we follow a consistent set of rules:
- Human in the loop – the systems we design prepare information, flag issues and make suggestions, while decisions remain with people. If the intended outcome of a project requires automated decisions about the situation of natural persons – for example concerning employment, access to a service or financial terms – before implementation we agree with the client on the requirements arising from the AI Act and from data protection law, and on the division of responsibilities between the parties.
- Lines we do not cross – we do not design or implement solutions covered by the prohibitions of the AI Act, in particular manipulative systems, social scoring or emotion recognition in the workplace or in educational institutions.
- Use-case assessment before implementation – we assess the planned use against the requirements of the AI Act and discuss the resulting obligations with the client.
- Disclosing interaction with AI – in solutions where a user talks to an AI system (a chatbot, voice assistant or conversational form), we design clear information about this fact, provided no later than at the time of the first interaction.
- Roles and responsibility – before implementation we agree with the client who acts as the provider and who acts as the deployer of the AI system, and what this division means for each party’s obligations. We do not provide legal advice – the final legal assessment is carried out by the client, independently or with the client’s own advisers.
- Transparency towards the client – the client knows in which parts of the solution AI operates, what type of models the system uses and what the limitations of those models are.
4. Recruitment and processes concerning people
In recruitment processes, AI tools play a supporting and preparatory role: they help organise submitted documents and prepare material for the person running the recruitment. The output of these tools does not replace a human reading of the documents, and no application is ever rejected automatically. Every application is assessed by a human, and personnel decisions – to invite a candidate to an interview, to hire or to decline – are made exclusively by people. We inform candidates about the use of AI tools in the recruitment process. A candidate has the right to receive an explanation of the role the AI tool played in the evaluation process.
We do not use tools that analyse biometric data, facial expressions, tone of voice or the emotions of candidates and employees. Any analysis of conversations and meetings concerns their substantive content only.
5. Human control
Materials prepared with AI support that we publish or hand over to clients are reviewed, verified and approved by a human. Responsibility for the content, meaning, accuracy and publication of the material rests with a human.
We are aware that AI tools can produce incomplete or incorrect content (so-called hallucinations), which is why their output is not published or passed on to clients without prior verification. In internal processes where AI organises or classifies information, a human always has the ability to inspect and correct how the tool operates.
6. Data protection and confidentiality
We apply the principles of caution and data minimisation when using AI tools:
- we use paid, business-grade versions of tools with training on our data disabled, wherever the provider makes this possible;
- we do not enter confidential data, sensitive data or information covered by trade secrets – ours or our clients’ – into AI tools available in free or consumer versions;
- if work on such information requires AI support, we use business-grade tools, on terms consistent with the agreement with the client, after a prior risk assessment and with appropriate safeguards in place;
- in supporting processes we use general, anonymised or minimised data;
- data entrusted to us by clients within projects is processed in accordance with the agreements concluded and with data protection law.
- before publishing or distributing AI-generated content, we verify that it does not infringe on third-party copyrights, trademark rights, or other industrial and intellectual property rights.
7. Security and limitations
We do not allow uncontrolled use of AI tools in a way that could put the interests of clients, business partners or the company itself at risk. Using AI does not release anyone from the duty of due care, confidentiality and compliance with applicable law and the agreed terms of cooperation.
The JAAQOB Holding team works under internal rules for the use of AI tools, which define permitted uses, data protection rules and the way irregularities are reported. We take measures to support the development of AI literacy among the people who use these tools on our behalf, in line with Article 4 of the AI Act – this includes an introduction to our rules for using AI tools and ongoing internal training.
8. Responsibility and oversight
Oversight of the use of AI tools at JAAQOB Holding is exercised by a person designated by the management board in our internal rules for the use of AI tools. Responsibility for the final result always rests with a human, not with the tool.
9. Disclosing the use of AI
Our labelling rules differ depending on the type of content, in line with Article 50 of the AI Act.
Text. Articles, posts and informational materials that have undergone substantive and editorial review by a human, and for whose publication a specific person holds editorial responsibility, are treated as authored content and are not labelled separately.
Images, video and audio recordings. If we publish material generated or modified by AI that depicts people, places or events in a way that could be perceived by the audience as authentic, we disclose that it has been artificially generated or manipulated. The disclosure is placed in the material itself, in a way that is visible or audible from the start. This does not apply to standard technical processing that does not change the message – for example corrections of lighting, framing or noise. Supporting graphics may additionally carry the caption “JAAQOB Holding | AI-assisted design”.
Chatbots and voice assistants. If we make available a solution in which a user talks to an AI system, we disclose this no later than at the time of the first interaction.
This notice describes our principles and does not replace the labels placed on specific materials or the messages displayed in the interfaces of our solutions.
Machine-readable labels. Where technically feasible and required by the AI Act, machine-readable technical metadata and watermarks are also used.
10. External documentation
Information about the privacy and security practices of the providers of the tools we use is usually available in their official documentation, terms of service and privacy policies.
11. The risk of uncontrolled AI use (shadow AI)
We are aware of the risks involved in the uncontrolled use of AI tools – in particular, situations where such tools could be used to process information without adequate control, safeguards or awareness of the consequences. We therefore follow the principle of conscious and controlled use of AI tools, and good practices in this area are part of the onboarding of every person joining our team, as well as of our regular contractors.
The use of AI at JAAQOB Holding is supportive in nature and does not replace personal professional responsibility, expert knowledge or independent human judgement.
This notice is descriptive. It is not a document required by the legislation on artificial intelligence, and it does not replace our internal procedures or the provisions of agreements with clients, which set out the detailed terms of cooperation.
Version 1.0, effective from: 2 August 2026. Previous versions are available on request.
The Management Board of JAAQOB Holding