Data residency
Data residency is a control that keeps a customer's data stored, and often processed, in a chosen geographic region such as the EU, so a team can meet its own legal, contractual or customer rules on where data lives.
Also called: EU hosting, regional processing, inference geo
By Divya Sharma · Checked Nov 15, 2026 · How we review
When you send prompts to a hosted model, the text travels to a data centre, gets processed, and may be stored for logs, batches or files. Data residency lets you pick the region where that happens. Teams ask for it when a customer contract, a regulator or an internal policy says data must stay in a given place.
Storage and processing are separate
Providers usually split residency into two parts. Storage at rest covers files, batch jobs and other saved data. Processing covers where the model actually runs on your request. OpenAI sets residency per project and notes that system data, such as account details, can sit outside the chosen region; it also charges a 10% uplift on data residency endpoints for models released on or after March 5, 2026. Anthropic has an inference geo setting per request and a separate workspace geo for stored data. Google Cloud says its global endpoints can process data in any of its locations and so give no residency guarantee.
What it does not cover
Residency is about place, not about how long data is kept. Pair it with zero data retention if you also need prompts not to be stored. It also does not cover tools, plugins or other third parties your app calls. If you need full control over location, self-hosting a model or using a regional inference provider are the alternatives. For whether a setup meets a specific law, read the provider's terms and get qualified advice; an audit report such as SOC 2 is a separate question.
Related terms
- Zero data retention
- Zero data retention (ZDR) is an arrangement in which an AI provider does not keep a customer's prompts and outputs after a request is answered, usually offered only to approved customers and only for eligible API features.
- Self-hosting (AI models)
- Self-hosting an AI model means running it on hardware you control, such as your own servers, rented GPUs or a laptop, instead of calling a provider's API. You choose the model and keep the data, but you run the infrastructure.
- Inference provider
- An inference provider is a company that runs AI models on its own hardware and sells access to them through an API, usually charged per token. You send prompts and get outputs without managing GPUs yourself.
- SOC 2
- SOC 2 is an audit report, produced by an outside CPA audit firm under AICPA standards, on a service organisation's controls relevant to security, availability, processing integrity, confidentiality or privacy.
- EU AI Act
- The EU AI Act, Regulation (EU) 2024/1689, is the European Union's law on artificial intelligence. It sets risk-based rules for those who build and deploy AI systems, from outright bans to transparency duties, phased in from 2025.