Google Cloud is offering customers using its generative AI services the chance to store data in their local region. It is a move by the hyperscale cloud provider to offer greater data residency options to its clients, following in the footsteps of its larger rivals, Amazon’s AWS and Google Cloud.
Data sovereignty has become a key issue for cloud providers in recent months amid fears that data transfers between the US and Europe may not be legal under GDPR. Alongside this, the rise of generative AI systems means tech leaders must consider how data used to train artificial intelligence models can be stored securely to protect user privacy and ensure proprietary information isn’t exposed.
Google Cloud launches new Vertex AI data residency regions
Google Cloud is attempting to address these two problems through ten new data residency regions for its Vertex AI services, which will allow users of the service to store their data in their location of choice.
A dedicated UK region has been set up for Generative AI on Vertex AI, which includes Google’s PaLM 2, Codey, and Imagen models, as well as the text embeddings and multimodal embeddings APIs. The US, Canada, the Netherlands, France, Germany, Belgium, Japan, Singapore and South Korea are other available regions.
Generalised North America and EU regions have been set up for the Vertex AI Search and Vertex AI Conversation tools.
Google Cloud plans to bring generative AI capabilities with data residency controls to customers in other parts of the world said Warren Barkley, senior director for product management and cloud AI at Google Cloud.
Barkley said: “Our new data residency guarantees continue our commitment to providing AI services with enterprise-grade scale, safety, security and privacy. Protecting customer data and privacy has always been core to Google Cloud’s mission, and this commitment is more important than ever as AI continues to evolve and mass adoption increases.”
Japanese bank Minna Bank is among the first businesses to utilise the new data residency regions. “This was a very important update for the financial industry and other industries that demand high standards of privacy, security, and governance because we can now control where data is stored,” said Masaaki Miyamoto, Minna Bank’s CIO.
Hyperscalers go big on data sovereignty
The new data residency regions are available now, Google Cloud says. The company is the third biggest player in the public cloud, albeit with a much smaller share of the market than its Big Tech rivals AWS and Azure.
Data sovereignty options are in high demand among cloud users in the public sector, as well as highly regulated industries like financial services. This is a particularly salient issue in Europe, as it is feared storing information about EU citizens on servers elsewhere may not be compliant with the bloc's GDPR data laws.
Transferring European data to servers in the US could be problematic for the likes of Google Cloud because US law allows government agencies to requisition data from tech companies on security grounds, something which is not compatible with GDPR. Previous transatlantic data-sharing arrangements to allow such transfers have been struck down in the courts after being challenged by privacy campaigners. A new pact, the EU-US Data Privacy Framework, was reached earlier this year, but it has yet to be tested legally.
To get around this, providers have been striking deals with European tech companies to create sovereign clouds - Google and Thales created a sovereign cloud for France in 2021. Microsoft Azure launched a sovereign cloud for Europe last year, and last month AWS launched a similar offering for customers.