Is a sovereign cloud needed for AI workloads?

 Asked if they are planning to use cloud to build AI solutions, 17% of organizations in Europe said they are already doing so, according to IDC’s 2023 EMEA Cloud Survey.

However, if you compare that number to those who are currently using sovereign clouds, the number is much higher — 42%. Why is that?

AI should be assessed the same way as all other apps and workloads destined for sovereign cloud (or indeed any cloud).

In essence, the need for sovereignty is typically mandated by the need to comply with regulations and data protection laws. But the top reason for those using sovereign cloud to build AI solutions is because of expanding cloud use.

Is your organization using or planning to use cloud to build AI solutions (sovereign cloud users)?

Source: IDC’s 2023 EMEA Cloud Survey (August 2023); n = 268

This is not a unique driver — it is one of the main reasons given by all organizations in Europe using or planning to use sovereign cloud solutions.

As cloud usage expands, many companies are factoring all aspects of security into their digital transformation agendas, including the possible need for data and cloud sovereignty.

Organizations should begin the process of implementing sovereign solutions by reviewing and classifying their apps and workloads according to how sensitive they consider the underlying data to be.

IDC research shows that the top 3 use cases that organizations currently or plan to use AI from a sovereign cloud include: process optimization; advanced data analytics (data classification, data governance, data mining, and insights extraction); and automated document processing (data extraction from human-readable documents).

How are these AI use cases expected to impact an organization’s sovereign cloud usage?

For most, the main issues identified include the possibility of new cybersecurity threats, the need to build additional knowledge, and increased cloud costs.

Security is again cited as the top challenge when sovereign cloud users are specifically asked about their main concerns about using AI in the cloud.

What are your organization’s main concerns when planning to use or using AI in the cloud (sovereign cloud users)?

Source: IDC’s 2023 EMEA Cloud Survey (August 2023); n = 426

The added protections sovereignty can bring should help secure all data for AI use cases deemed to be sensitive.

The need to bolster cybersecurity therefore emerges as the main pain point providers will need to address for organizations using sovereign cloud to build their AI solutions.

IDC believes cybersecurity is a shared responsibility between vendors and their clients: The IT provider has responsibility for security of the cloud, while customers ensure they remain secure in the cloud.

This is discussed in the next blog in this series, “Implementing Digital Sovereignty: The Decision Framework.”