Legal

Sub-processors

The companies that process data on Filo's behalf, what each does, and where.

Effective October 21, 2026Version 0.1

This is a draft. It takes effect when Filo signs its first customer agreement, after review by our lawyers. Until then, it describes what we intend and how Filo works today.

What a sub-processor is

A sub-processor is a company Filo uses to run the service, and that handles your data to do so. This page lists each one, what Filo uses it for, and where it runs. Systems you connect yourself, such as your Google Workspace, your Slack or your Fiddle account, are not sub-processors: you hold those accounts, and Filo reaches them only as you.

The list

Provider What for Where
Google Cloud Hosting, storage and networking for Filo (Kubernetes, object storage), and the servers Filo's database runs on United States, us-central1
Resend Account email and the acknowledgements Filo sends, from filoagent.com United States
Anthropic Models that read, draft and reason in Filo. Content is sent under Anthropic's commercial API terms; it is not used to train models. United States
OpenAI, Google (Gemini), DeepSeek, OpenRouter Models, only when a platform admin picks that provider for a workspace. Same terms: API use, no training on your content. United States, except DeepSeek (China)

Filo sends a model only what a turn needs: the messages in the chat, the files attached to it, and the records the agent reads for that turn.

Changes

New sub-processors are added to this page with the date, and workspace admins are emailed 30 days before a new one handles customer data. Questions go to privacy@filoagent.com.