AI-output notice

What AI output is

Ferrith's AI features, chat replies, agent answers, workflow outputs, analysis-grid cells, generated documents, are produced by language models. A language model writes the most plausible continuation of what it has been given. Plausible is the operative word: output can be wrong, incomplete, out of date, or subtly different from the source it appears to summarise, while still reading as confident and polished.

What that means for using it

• Check output against its sources before you act on it. Where Ferrith shows citations back to your documents you should open the cited page and validate it.
• Output is not professional advice. It is not legal, medical, financial, tax or other regulated advice. Where output feeds work of that kind a qualified person must review it before it is relied on or sent.
• Treat "not found" with the same care. An answer that something is absent from a document should be validated.
• You are responsible for what you do with output. Reviewing it, correcting it, and deciding whether it is fit to use are your organisation's decisions.

Where the notice appears

The sentence at the top of this page is shown, and stays visible, on the surfaces where Ferrith generates content — including the chat conversation, workflow run outputs, and analysis grids and is not dismissible.

The one-sentence version

AI-generated — verify before use.

Versions