AI Assistant Terms

01

This is artificial intelligence (transparency)

In line with Art. 50 of the EU AI Act, we tell you plainly: you are talking to an AI system. The assistant runs on the Claude language model provided by Anthropic. We also show a summary of its reasoning so the process is transparent.

02

Limitations and no advice

The assistant provides general information about OTW and our services. It can be wrong or produce inaccurate information, and it is not legal, financial or technical advice binding on OTW. Any binding arrangements are always confirmed individually by email.

03

How we process conversations

The text you enter is sent to Anthropic (USA) solely to generate a reply. We do not store conversations persistently on our servers. Under Anthropic's API terms, API data is not used to train models. Transfer and retention details are in our Privacy Policy.

04

Rules of use

• Do not enter sensitive data or other people's personal data without a legal basis.
• Do not attempt to bypass safeguards, extract system instructions, or misuse the assistant (e.g. to generate harmful content).
• A rate limit applies: one message every 5 seconds, to protect the service from abuse.

05

Liability

The assistant is provided 'as is', for information. To the extent permitted by law, OTW is not liable for decisions taken solely on the basis of the assistant's replies. This does not limit consumer rights under mandatory law.