How will Apple Intelligence decide when to offload a query to ChatGPT?

Apple Intelligence decides when to offload a query to ChatGPT based on the complexity of the task and the capabilities required to handle it. For simpler tasks, Apple Intelligence uses its on-device foundation model, which is optimized for performance and efficiency. However, for more advanced features that require larger models to reason over more complex data, Apple Intelligence extends to the cloud using private cloud compute. This allows it to leverage larger foundation models while maintaining user privacy and security.

For more details, you can refer to the Platforms State of the Union session.