Transparency & AI
Last updated: April 2026
How we use artificial intelligence — and where we choose not to.
Our Commitment
At .WORKS, we believe that the best way to use artificial intelligence is to be open about it. To specify where, how, why — and above all, where it should do nothing. This page exists so our clients know exactly what to expect, well before the European AI Act officially requires us to do so in August 2026.
We see AI as just another tool in our workshop. A good tool when guided by a human hand; a bad choice when it replaces what should remain alive, sensitive, and authentically crafted.
What we use — and why
In our daily workflow, certain AI-assisted tools help us gain in precision and quality. The main ones are:
- Claude Code (Anthropic) — assistance with technical writing, code structuring, and reviewing automation configurations.
- Gemini (Google) — targeted research, synthesis of technical documentation, and cross-checking.
They are involved in well-defined tasks: accelerating a search, clarifying a structure, proofreading a technical paragraph, generating a code or automation flow skeleton that we then modify by hand.
What remains entirely human
Certain parts of our work never go through generative AI, because they constitute the very essence of what we do:
- Graphic creations based on sacred geometry are drawn and composed manually by Patricia Faucon. No generative image AI is involved in their design. Each piece is an original human creation, copyrightable for the benefit of the client.
- Art direction and aesthetic choices are driven by a human vision, nourished by the study of symbolic traditions, geometry, and a sensitive approach to form.
- Client relationships — listening, advising, questioning, decision-making — are handled directly by Patricia Faucon, without any automated intermediary.
Supervision and validation
Every deliverable that leaves the studio goes through human validation. Nothing is delivered “raw” from an automated tool. This supervision is our minimum commitment: what we sign off on, we have read, understood, and validated.
When an AI tool has played a significant role in a deliverable — for example, generated codebase, pre-written technical text, or a suggested automation structure — we explicitly state this to the client in the corresponding delivery document.
You, as a client
If we build a system for you that contains or uses artificial intelligence (chatbot, intelligent automation tool, assistance feature), you become, under the AI Act, a deployer of that system. This entails a few simple but important obligations for you:
- inform your end-users when they interact with an AI;
- maintain human oversight over automated decisions;
- keep a record of potential incidents.
Upon delivery, we provide you with the necessary documentation to fulfill these obligations without difficulty. If you wish, we can also assist you in training your teams (an “AI literacy” obligation in effect since February 2025).
Your data
The AI tools we use daily are never fed with your confidential data without your explicit consent. Practically:
- the sensitive contents of your projects (strategic documents, personal data, trade secrets) are handled with tools whose privacy policies guarantee no training on user data;
- the professional accounts we use (Claude, Gemini) are configured to disable learning on our conversations where this option exists;
- when in doubt, we ask you first.
Accessibility
Technological transparency goes hand in hand, in our view, with digital accessibility. A site that lies about its tools and a site that excludes people with disabilities stem from the same indifference.
.WORKS designs its web deliverables in compliance with WCAG 2.1 level AA standards, consistent with the European Directive 2019/882 (European Accessibility Act), transposed into Belgian law by the law of November 5, 2023. Concretely, this means: keyboard navigation, screen reader compatibility, sufficient contrast, alternative text on images, and properly labeled forms.
When relevant to your business, we also integrate an accessibility statement on your sites, so your visitors know what to expect and how to report an issue.
Legal framework
This page falls under:
- Regulation (EU) 2024/1689 (AI Act), particularly its Article 50 on transparency obligations, applicable from August 2, 2026;
- Directive (EU) 2019/882 (European Accessibility Act), transposed into Belgian law by the law of November 5, 2023;
- Regulation (EU) 2016/679 (GDPR) for all matters regarding personal data.
Our full commitments regarding data protection are described in our privacy policy.
Questions?
If you have any questions about our practices, our use of a particular tool, or how we could adapt our methods to your own compliance requirements, simply write to us at patricia@faucon.works.
This page is reviewed at least once a year, and upon each significant evolution of our tools or the regulatory framework.
