What changed

A new buyer’s guide released by theCUBE Research highlights the complexities of AI sovereignty, focusing on control over data and models. The guide outlines five pillars—territorial, operational, technological, legal, and financial—to assess how well providers manage data sovereignty. Amit Eyal Govrin, lead analyst for sovereign AI at theCUBE Research, emphasized that sovereignty is not a single checkbox but a stack of control layers that buyers must examine.

What this means in practice

The guide evaluates seven hyperscalers across 40 offerings, grading them based on 17 yes-or-no tests. Legal aspects are assessed separately due to the influence of local laws. Govrin noted that while Google’s air-gapped offerings scored relatively well, no hyperscaler had a clear answer to the U.S. CLOUD Act. This means that if a U.S. court issues a production order, data stored with a hyperscaler could be compelled to be turned over, effectively making it U.S. data.

Limitations and availability

The guide’s findings highlight the lack of clarity around the U.S. CLOUD Act’s impact on data sovereignty. Govrin pointed out that even if data is hosted in a sovereign region, such as AWS Frankfurt for German customers, there is no guarantee that data will remain under the control of the data owner. The guide also underscores the importance of legal frameworks in shaping AI infrastructure choices, with the U.S. CLOUD Act and incoming European regulations playing a significant role. The supplied announcement does not specify the exact scoring methodology or the full list of hyperscalers evaluated.

Source facts for AI Sovereignty Guide Highlights Control Concerns in Data Centers: A new buyer’s guide sets out to measure that control provider by provider. And the first one is, are you in fact comfortable … ceding that control to somebody else? Govrin spoke with theCUBE Research’s John Furrier at SC26, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio.
Original AINewsTrend explainer based on the cited source. This graphic summarizes source statements; it is not an independent benchmark.

Sources and further reading

  1. siliconangle: Enterprises fear AI success more than failure, because success means lock-in they can’t escape