We use OpenAI’s and Anthropic’s language models. This means that text content within your uploads may, in unmarked chunks, be submitted to OpenAI or Anthropic as context when generating a response. We will never send your documents in its complete original form to any other organization, including OpenAI, Anthropic or their partners.

To better explain the associated implications, we like to use the analogy of mixing bags of legos (chunks of documents) and sending only a mix of blocks to OpenAI at a time. In the right combination, your blocks will form your original document (like a lego castle). However, when sending to OpenAI, your blocks can come from different documents, and they are sent alongside thousands of blocks from other users. There is basically no way (and no material benefit) for OpenAI to attempt reconstructing your blocks and learn of your original documents’ content. OpenAI has also explicitly stated that it will not use data submitted by customers via its API endpoints to train or improve models (https://openai.com/policies/api-data-usage-policies).