No. Your data is not used to train any AI anywhere.
Unlike many other AI scribes in the market that collect and use your data to improve their models, Klarify maintains a strict no-training policy. Your sensitive therapeutic conversations remain private and are:
Stored in Canada
Never sold or shared for anyone else's benefit
Never used to train or fine-tune any AI models
Starting November 1, 2026, Klarify may use de-identified, aggregated information, such as which features get used most, to make the product better. It cannot identify you or your clients, and it is never used to train AI.
We've specifically designed our infrastructure to prioritize your privacy and that of your clients. This commitment to data protection is what sets Klarify apart in the AI therapy tools space.
Which AI models does Klarify use?
Klarify's AI features use models made by Anthropic. We run them through AWS Bedrock, so your requests are handled by AWS and are not sent to Anthropic's own servers. Neither AWS nor Anthropic can use your data to train models. By default, processing happens in Canada or the US. You can switch to European processing in your settings. For details, see Security & Privacy at Klarify.