How we use artificial intelligence

As support in engineering, design and documentation: exploring alternatives, writing drafts, reviewing code, preparing material. Never as a substitute for professional judgement or for our responsibility for what we deliver.

It is not a separate service category or a selling point. It is one more tool, and we use it where it helps.

No code, text or decision generated with the help of AI goes into production, or to you, until a person has reviewed it and taken responsibility for it. If something fails, we are responsible, whether or not a tool was involved.

We do not enter your confidential information into tools that do not contractually guarantee that they will not use it to train their models.

Nor do we paste credentials, keys or your users’ personal data into them when the work does not need them. The practical rule is simple: if a piece of data is not needed to solve the problem, it stays where it is.

When a product we develop incorporates artificial intelligence that interacts with people —a conversational assistant, for example— or that generates synthetic content, we make this clear to users. Article 50 of the EU AI Act, applicable from August 2026, requires this, and we also think it is the right thing to do.

We do not deploy systems that make decisions on their own with a legal or significant effect on a person where no human can intervene.

If an engagement falls within the high-risk categories of the AI Act (recruitment, lending, biometrics, education where it affects access, critical infrastructure), we say so before starting and handle it with specialist advice, not as routine development work.

You can ask for this and we will agree it in writing before starting. It usually affects the timeframe, and sometimes the price, because part of the work is done differently. Write to us at info@bazrox.com and we will discuss it with concrete figures, not in the abstract.

We do not certify anything. There is no seal or external audit here, only commitments that depend on how we work, and we list only those.

The providers of these tools may change their terms without notifying us. We choose service tiers that do not train on the data we submit, and we review this, but we do not control their terms.

Who holds the rights in what is generated with the help of AI is a legal question that is still unsettled in Europe. Our side is clear: what we develop specifically for your project is assigned to you by contract, once the corresponding amounts have been paid, in an express clause rather than taken for granted.

Write to us at info@bazrox.com and we will tell you which tools are involved in that engagement and with what guarantees. It is more useful than a general answer, because not every project uses the same thing.