How we use AI

1. People stay responsible

Conscious communication helps you clarify a goal, give useful context, ask questions and check whether an answer fits. These are useful skills for working with AI while keeping your own voice.

Our aim is to support human judgement and human connection. AI output is something to review, not a decision you must follow.

2. Where AI can be involved

• Satellite Scan: coaches prepare your dashboard. The website also has an enabled-provider pathway for generating dashboard visualisations from a prompt and supplied data. “Coach-prepared” does not mean every step is manual or that AI only summarises.
• Prompts and resources: AI can help draft example requests and supporting material. When you use a prompt in your own AI tool, that provider’s terms and data settings apply.
• Communication visualisations and content: Thesys can generate visualisations and drafts when the relevant connection is enabled. Its current dashboard request uses a model identified as Anthropic Claude through Thesys.
• Website illustrations: some refreshed website scenes were generated or edited with AI. They illustrate possible practice settings and are not evidence of real clients, staff or training sessions.
• Meeting notes: optional Fathom recording and AI notes require explicit agreement before recording starts; declining recording does not prevent the discovery call.

Tool availability depends on the service and its enabled settings. Ask which tools and data will be involved in your own service before sharing sensitive information.

3. Information sent to AI tools

For the website’s Thesys generation pathway, the request includes the prompt and any data supplied with it. That information is processed by Thesys and its model-provider arrangements.

Do not assume a consumer AI account is private or that a provider never uses inputs for training. Retention, training use, subprocessors and transfer safeguards depend on the applicable service agreement and settings. These need service-specific confirmation.

Use fictional or suitably anonymised practice examples where possible. Do not upload confidential client or colleague information without the authority to do so. See the privacy notice for recipients and rights.

4. Checking outputs

AI may invent facts, omit context or reproduce bias. Check names, dates, calculations, links and tone before using an output. Consider how the person receiving it may understand it.

Ask the coach what was generated, what was checked and what remains uncertain. This notice does not promise that every output has been labelled or reviewed, or that all analysis is human-only.

Communication reflections and ACX levels are learning tools, not clinical diagnoses or suitability scores for hiring, credit or other decisions affecting someone’s rights.

5. AI literacy and the law

The EU AI Act is a reference for reviewing responsible use. Our ACX map and learning offers do not certify legal compliance. An AI system’s legal risk classification depends on its purpose and use; this page does not classify every application as minimal or limited risk.

See the European Commission’s AI literacy guidance. Learning a framework does not replace reviewing the systems used in your organisation.

The teaching approach supports transparency, appropriate human oversight and data protection. It does not authorise manipulation, exploitation of vulnerabilities or automated decisions about people’s rights.

6. Your choices and concerns

• Ask [email protected] which AI tools are involved and how they contribute to an output.
• Ask whether a human-only option is available before booking or submitting data. This is a request to discuss alternatives, not a guaranteed option for every service.
• Question or correct an output and ask for a human explanation.
• Exercise applicable access, objection, consent and automated-decision rights under the GDPR. These depend on the processing and legal conditions; they are not limited to what is technically convenient.
• Complain to a relevant data protection authority. See the privacy rights section.

7. Coaching, interviews and retreats

The existing coaching, interview and retreat services keep their agreed scope. Practising clear goals, listening and thoughtful feedback can support AI literacy, but this connection does not turn those services into new AI offers.

AI training and coaching are booked according to the specific offer. Do not assume AI training is included in a retreat, interview bundle or communication session.

8. Research and updates

The ACX teaching map draws on Estève’s AI and communication research, including work described at Arbora research. Visit Arbora.partners for more about that project. Green Elephant and Arbora are separate projects; this reference does not mean your service data is shared with Arbora.

We may update this notice as tools and practices change. For questions or concerns, contact [email protected].