Highlights:

  • Data residency options are now offered for Vertex AI’s pre-trained generative models, including PaLM-2 and specialized neural networks for various tasks.
  • The rollout of a second wave of data residency controls is occurring for Vertex AI Search and Vertex AI Conversation.

Google LLC’s cloud division has recently unveiled additional data residency functionalities for Vertex AI, its set of managed services for artificial intelligence development.

Industries subject to regulations frequently need to ensure that their data remains within the geographical jurisdiction of its origin. Google Cloud’s newly introduced data residency features aim to fulfill this need. As per the company, users can designate the storage location for data processed by Vertex AI using these features.

The initial set of data residency choices is being introduced for Vertex AI’s pre-trained generative AI models suite. This collection encompasses PaLM-2, a general-purpose language model akin to GPT-4, and specialized neural networks optimized for code and image generation tasks. Vertex AI empowers companies to integrate these neural networks into their applications.

Google Cloud is now offering users of its generative AI models the ability to store processed data in one of ten countries: the U.S., Canada, Japan, Singapore, Korea, the Netherlands, France, the U.K., Germany, and Belgium. This same option will also be extended to Vertex AI’s embedding APIs, aiding developers in converting their datasets into a format that neural networks can more efficiently handle.

A second wave of data residency controls is being introduced for Vertex AI Search and Vertex AI Conversation. The former facilitates developers in integrating search functionalities into their software, while the latter streamlines the development of chatbots.

Users of these two services now have the option to choose Google Cloud as their data storage location, either in the U.S. or the European Union. Upon activating the new data residency features, Google maintains the data within the specified jurisdiction and conducts processing activities in that location.

In a blog post, Warren Barkley, Senior Director of product management for cloud AI at Google Cloud, stated, “This is just the start of our ongoing commitment to continuing to expand data residency for our customers.” He added that more offerings are on the way.

The rollout of these new features coincides with increased attention from Google’s competitors on data residency.

In the preceding month, Amazon Web Services Inc. launched the AWS European Sovereign Cloud, a forthcoming network of data centers in the EU. These facilities cater to highly regulated organizations needing to store their data within the bloc. AWS asserts that the European Sovereign Cloud will operate independently from its current data centers and will be overseen by employees residing in the EU.

Previously, Oracle Corp. launched a comparable solution named the EU Sovereign Cloud. Meanwhile, Microsoft Corp. is also actively investing in data residency. The company is unveiling a program that empowers customers using Microsoft 365, Azure, and other services to maintain their data within the EU.