AI transparency · updated 16 September 2026

How we use artificial intelligence, and what stays in human hands

We help companies use artificial intelligence, and we use it ourselves. Here we explain where, under what rules, and what stays in human hands. Anyone who writes to us should know who they are talking to. Anyone who works with us should know what happens to their documents.

Our rules in client work Where the data goes

Illustration showing the four places on the site where artificial intelligence is involved: the chat, WhatsApp messages, articles and illustrations, each with the way it is disclosed. In the centre, the rule: a person decides. — Illustration

AI disclosed: every assistant introduces itself as AI
A person decides: no choice about clients, candidates or employees is left to a system
No invented people: no generated faces, voices or scenes on the site

01 · copy the link to this sectionIn brief

Why this page

Anyone who brings artificial intelligence into companies should be the first to explain how they use it. The rules point the same way. In Europe, anyone who offers a system that converses with people must make that known. In Italy, anyone practising one of the ‘intellectual professions’ must tell clients which AI tools they use. We prefer to say it plainly, before anyone asks.

Disclosed assistants. On WhatsApp an AI assistant answers, and says so. The same applies to the site chat.
Articles. They may be produced with the help of AI. They are published under the responsibility of the Strategic editorial team.
Illustrations. They are drawn graphics. No people are generated.
Client work. AI helps with the preparation. People check, decide and are accountable.

02 · copy the link to this sectionOn the site

Where you will find it on this site

There are five places, and for each one we say how it is disclosed.

Example conversation. The assistant introduces itself as Strategic’s AI assistant, answers by citing the page of the site it takes the information from and, on request, arranges for a person to get back to you. Beside it, how it introduces itself and what not to put in the chat. — Example · the conversation shown is not real

The site chat The site chat is run by Maxim, Strategic’s AI assistant. It answers questions about our services based on the published pages and shows the page it takes the information from. It introduces itself as an AI assistant.

WhatsApp Messages to Strategic’s WhatsApp number are answered by an AI assistant, which introduces itself as such.

Observatory articles Some articles are prepared with the support of artificial intelligence tools, for research and the first draft. They are published under the responsibility of the Strategic editorial team, which gives the sources for the laws and data cited. In the editorial dashboard, an AI assistant can suggest the title and the description for search engines. The editorial team accepts or corrects them.

Illustrations and cover images The images on the site are illustrations drawn for the purpose. They show example screens and cards, with invented data and the label ‘Example’ wherever they might look real. Article cover images are put together by software from the title and the subject area. Artificial intelligence tools may also be used in preparing the graphics, but no image shows invented people, faces or events. Photographs, where there are any, are real photos used under licence.

Where you won’t find it The contact form does not go through an AI assistant. Your request reaches the people at Strategic. On the Blackroom page there is no chat and no assistant. You only ever write to and speak with people.

03 · copy the link to this sectionThe rules

In client work: the rules we set ourselves

Artificial intelligence helps us read documents, prepare drafts, write code and translate. It does not change who is accountable for the work.

We say so first, then we use it. In the proposal we state whether and where we will use artificial intelligence tools. If you would rather they were not used on your work, we put that in writing and allow for it in the timeline.
Your documents stay under control. Which tools may read an organisation’s documents, where the data resides and whether the vendor may use it to train its models are all decided with the organisation before work begins, and put in writing.
A person checks. Every document, analysis or line of code we deliver is checked by a member of the project team, who is accountable for it.
Decisions stay with people. No assessment of clients, candidates or employees is left to an automated system. AI prepares the material. A person decides.
Blackroom stays out of it. We do not use artificial intelligence tools on owners’ confidential matters.

The same rules, applied to your systems

Example diagram. A draft prepared by the AI assistant goes to a person for checking, then to the person in charge for a decision, and the log records who checked, who decided and when. The people on the team are accountable for the work.

The path of every deliverable

01AI preparesDrafts, summaries, comparisons, code.

02A person checksSources, data and accuracy.

03A person decidesThe person accountable on the project team.

04A record is keptOf who checked and who decided.

04 · copy the link to this sectionThe data

Where the data goes

Chat conversations. They are used to answer you and, if you ask, to get back to you. They are kept for the period stated in the privacy policy. The text of your questions is passed to the provider of the language model that generates the answers, OpenAI, within the limits described in the privacy policy.
WhatsApp messages. They pass through WhatsApp, through the servers of the services that handle the replies and through the providers of the language models used for the replies. These providers are listed in the service documentation, and you can ask for the current list through the contact form.
No profiles of the people who write to us. Conversations are not used to build advertising profiles, and they are not sold or passed on to others for their own purposes.
Your rights. You can ask to access, correct or delete your data. The privacy policy explains how.

05 · copy the link to this sectionReference

The rules that apply

Updated 16 September 2026

Law	What it requires, in brief	How we apply it
AI Act, Regulation (EU) 2024/1689, Article 50(1), in the European Union, from 2 August 2026	AI systems intended to interact directly with people must be designed so that the people using them know they are dealing with an AI system	The chat and the WhatsApp assistant introduce themselves as AI
AI Act, Article 50(4)	Anyone who publishes AI-generated text to inform the public on matters of public interest must disclose it, unless the text has undergone human review or editorial control and someone holds editorial responsibility for it. The same article requires disclosure of images, audio and video that realistically imitate people or events	Articles are published under the responsibility of the Strategic editorial team. No image imitates real people or events
Law No. 132 of 23 September 2025 (Italy’s AI law), Article 13, in Italy	In the intellectual professions, AI is only a support tool and intellectual work must remain predominant. Information on the AI systems used is given to the client in clear, simple and comprehensive language	In our proposals we state which AI tools we will use
GDPR, Regulation (EU) 2016/679, Article 22	The right not to be subject to decisions based solely on automated processing that produce legal effects or similarly significant effects	No decision about people is taken by an automated system
Law No. 171 of 21 December 2018, in San Marino	Protection of individuals with regard to the processing of personal data	The processing carried out on the site is described in the privacy policy

This is general information and does not constitute legal advice.

Sources (consulted on 16 September 2026)

06 · copy the link to this sectionPrinciples

What we don’t do with artificial intelligence

We don’t pass off an AI assistant as a person.
We don’t generate faces, voices or people for the site or for our communications.
We don’t leave it to a system to decide about clients, candidates or employees.
We don’t put client documents into tools that have not been agreed.
We don’t use AI assistants in Blackroom.

Artificial intelligence prepares. People decide, and they are accountable.

07 · copy the link to this sectionFor businesses

The same rules, on your systems

Example of a register of the artificial intelligence systems in use in a company.

AI AI security and governance Usage rules, a register of systems and testing of the assistants that read company data.

AI AI training People who use the tools learn to read an answer and check its source.

Example of Company Brain: documents and data linked to an answer with its sources.

AI Company Brain An assistant on the company’s own documents that shows where each answer comes from.

08 · copy the link to this sectionFrequently asked questions

Frequently asked questions about how we use artificial intelligence

Am I talking to a person or to an AI?

On WhatsApp an AI assistant answers, and says so at the start of the conversation. The same applies to the site chat. The contact form, on the other hand, goes straight to the people at Strategic.

What happens to what I write in the chat?

It is used to answer you and, if you ask, to get back to you. The text is passed to the provider of the language model that generates the answers. It is kept for the period stated in the privacy policy, which also explains how to ask for it to be deleted. Please do not put confidential data or company documents in the chat.

Are the articles written by artificial intelligence?

Some are prepared with its help, for research and the first draft. They are published under the responsibility of the Strategic editorial team, which gives the sources for the laws and data cited.

Are the images generated with artificial intelligence?

The images on the site are illustrations. They are example screens and cards drawn for the purpose, with invented data labelled as an example. Artificial intelligence tools may also be used in preparing them, but none shows invented people, faces or events.

Do you use artificial intelligence on our company’s documents?

Only if it has been agreed. Before work begins, we set down in writing which tools may read the documents, where the data resides and whether the vendor may use it to train its models. Everything we deliver is checked by a member of the project team.

Can we ask you not to use it on work that concerns us?

Yes. Tell us before work begins and we will write it into the proposal. Some tasks may take longer, and the proposal allows for that.

Who updates this page?

Strategic’s management updates it. The page shows the date of the last update. It changes when the tools we use, or the rules that apply to them, change.

09 · copy the link to this sectionContact

A question about how we use artificial intelligence?

Write to us through the contact form. For requests about your personal data, the privacy policy explains what to do. We reply within two working days.