European companies can store data from generative AI models locally thanks to the new Google Cloud offering. The option is available for the Vertex AI portfolio and the Vertex AI Search and Vertex AI Conversation tools.
With the introduction of data residency options for its Vertex AI portfolio, Google Cloud wants to make it easier for companies to be able to store their (generative) AI data in the jurisdictions in which it was generated. This option helps comply with local laws and regulations.
The new options include those from the company’s own PaLM-2 models and those from more specialized Codey, Imagen models and the Text Embedding and Multimodal Embedding APIs.
Data storage in different locations in Europe
The data from these specific LLM models can now be optionally stored in a dozen countries: the US, Canada, Japan, Singapore, South Korea, the Netherlands, France, the UK and Belgium. Not only does the tech giant store the (generative) AI data in these countries, but the gen AI data can also be generated in the designated areas.
Also for Vertex AI Search and Conversation
In addition, the new data residency options and features also apply to the Vertex AI Search and Vertex AI Conversation tools. The first tool enables search functionality to be built into software, and the second should simplify the development of chatbots.
For these two tools, users can now choose whether to store their data in the European Union or the U.S. Customers can store their data within multiple regions in this umbrella location for both options.
Data residency initiatives from other parties
Google Cloud is not the only public cloud environment or tech giant that allows customers to process and store their data in specific regions or countries with data residency options. AWS recently introduced a dedicated data center network with AWS European Sovereign Cloud.
Oracle already introduced a similar EU data residency solution last summer, and Microsoft is currently working on its own program to house Microsoft 365 data in specific EU environments.