Legal
We use artificial intelligence to power the coaching features in Throughline. This page explains what the AI does, what content is sent to it, and what that means for you.
The AI analyses the structure and clarity of your communication drafts and suggests rewrites. Where you have supplied a source document, it also checks whether the figures and names in a draft appear in that source, and flags the ones that do not. It does not verify anything against the outside world, and it does not provide legal, medical, or financial advice.
We use Claude, a large language model developed by Anthropic, Inc., accessed via their commercial API.
The content you put into Throughline: the stories you write, any source document you upload, any text you paste in for review, and any grant application material you draft. Governance policies you upload are sent only when you run a compliance check against them. It is sent as written, with no redaction step, so any name or figure your draft contains goes with it. We do not send your account name, email address, or payment information to Anthropic.
Under Anthropic's API terms, content submitted through the API is not used to train their models by default.
AI outputs can be compelling and wrong at the same time. You are responsible for everything you publish.
Their frameworks are built into the AI's instructions, but responses are generated by a language model, not by them personally.
Contact phil@cornerstonemedia.com.au with any questions about how AI is used.