AI use statement

Mule uses AI tools to work faster. This statement says where, under what terms, and what never happens.

Where we use AI

• Drafting replies, summaries of long message threads and first drafts of proposals and documents.
• Helping write and review code.
• Classifying and routing incoming briefs so the right person sees them first.

Rules we follow

1. A person reviews everything before it is sent to you or shipped. No AI tool sends you a message on its own.
2. Commercial terms only. We use AI providers under business terms that do not allow them to train their models on your data. We do not paste client personal data into consumer AI tools.
3. No automated decisions about people. AI output is never the sole basis of a decision with legal or similarly significant effects about an individual.
4. Disclosure. Any chat or assistant feature on our sites or in systems we build says clearly that you are talking to an AI system, as the EU AI Act requires.
5. Staff training. People at Mule who use AI tools are trained in their limits and risks.

In the systems we build

Where a client system includes an AI feature, it is documented in the project's specification, it is labelled in the interface, and the client decides whether it is on. Usage is metered and shown in the Portal like any other service.

The Mule tag

The tag ranks animals worth a closer look. It does not diagnose. People decide on animal care.

Provider

Our AI provider is listed in the Sub-processor List. Questions: juastins@icloud.com.