Data retention policy
Data will not be stored for more than the amount of time specified in our Terms of Service, Privacy Policy, customer contracts, or other binding agreements.
Data archiving and removal policy
By default, a customer’s data is stored indefinitely. Upon customer request, the data may be deleted up to 1 month after the request, at the latest, with the exception of data that is required to establish proof of a right or a contract, which will be stored for the duration provided by enforceable law. Once deleted, a user’s data cannot be restored.
Data storage policy
Assembled’s data is encrypted at rest using AES-256.
App/service has sub-processors
yes
Guidelines for sub-processors
App/service uses large language models (LLM)
yes
LLM model(s) used
OpenAI GPT-4 (via ChatGPT Enterprise)
LLM retention settings
OpenAI retains API data for up to 30 days for abuse and misuse monitoring, but does not use the data for training or improvement of models. We do not store or log user prompts or responses beyond what is required for operational performance.
LLM data tenancy policy
Each customer’s data is isolated and processed separately. As a ChatGPT Enterprise customer, our queries are not used to train OpenAI models. All API usage is tenant-specific and not shared across customers.
LLM data residency policy
As of now, OpenAI processes and stores data in the United States. (Note: OpenAI does not currently support custom regional data residency, but this may change—double-check with OpenAI if data residency is a contractual requirement.)