Security & Privacy
The security measures for the GRID API service in its current Alpha phase are managed in the same way as those for GRID’s Calculator Studio product (see details here).
It's important to note that most use cases for the GRID API do not involve spreadsheets containing personally identifiable information (PII). Instead, these use cases involve transforming template spreadsheets into RESTful API services, allowing calculations to be performed by passing input values through an API call and returning the results. Even if input values originate from a source containing PII, these values are not tied to any specific record within the API call or the spreadsheet that performs the calculations.
If a use case requires including PII in either the spreadsheet or the request data, please refer to the Calculator Studio security documentation.
In the context of using large language models (LLMs), it's worth noting that while the GRID API may be called through function calls (also known as "tool use" or "actions") from your LLM model, all chat, text, or file data—apart from the spreadsheet model itself—is managed on the LLM side in compliance with its own security policies. The GRID API service does not use request data, response data, or spreadsheet files for training any LLM models.