At IBM Think 2026, Priya Srinivasan of IBM, general manager of IBM Software, discusses digital and artificial intelligence sovereignty and the newly released Sovereign Core platform. The conversation examines the evolution from data sovereignty to AI sovereignty, the software-first hybrid deployment model, Red Hat OpenShift integration, and the partner ecosystem that supports enterprise governance and hybrid cloud strategies.
Srinivasan explains that Sovereign Core is generally available and that the software stack places the control plane inside the customer boundary to ensure operational resiliency and what they call "sovereignty with receipts." They emphasize no vendor lock-in, hardware-agnostic deployment, an extensible catalog with golden patterns, turnkey options through partners and a consumption-based resource unit pricing model. The discussion also highlights a managed service provider go-to-market focus and the goal of accelerating AI initiatives from months to hours.
Watch this discussion to learn how Sovereign Core addresses digital sovereignty, AI governance, cloud native deployment and operational resilience while enabling enterprise artificial intelligence at scale.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
IBM Think 2026. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Register for IBM Think 2026
Please fill out the information below. You will receive an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for IBM Think 2026.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
IBM Think 2026. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Sign in to gain access to IBM Think 2026
Please sign in with LinkedIn to continue to IBM Think 2026. Signing in with LinkedIn ensures a professional environment.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
Sripriya Srinivasan, IBM
At IBM Think 2026, Priya Srinivasan of IBM, general manager of IBM Software, discusses digital and artificial intelligence sovereignty and the newly released Sovereign Core platform. The conversation examines the evolution from data sovereignty to AI sovereignty, the software-first hybrid deployment model, Red Hat OpenShift integration, and the partner ecosystem that supports enterprise governance and hybrid cloud strategies.
Srinivasan explains that Sovereign Core is generally available and that the software stack places the control plane inside the customer boundary to ensure operational resiliency and what they call "sovereignty with receipts." They emphasize no vendor lock-in, hardware-agnostic deployment, an extensible catalog with golden patterns, turnkey options through partners and a consumption-based resource unit pricing model. The discussion also highlights a managed service provider go-to-market focus and the goal of accelerating AI initiatives from months to hours.
Watch this discussion to learn how Sovereign Core addresses digital sovereignty, AI governance, cloud native deployment and operational resilience while enabling enterprise artificial intelligence at scale.
>> Digital sovereignty and AI sovereignty is one of the hottest topics that we hear about on theCUBE as we go around the world. My name is Dave Vellante, and I'm with theCUBE Research and SiliconANGLE and theCUBE. And we're here at IBM Think 2026 in this special studio that IBM has built for us. I appreciate that, IBM. And I'm here with Priya Srinivasan, who's the general manager of IBM software. And we're going to talk about sovereignty. Priya, thanks so much for spending some time with me.>> Of course. Thank you.
Dave Vellante
>> So as I said in my open, everywhere I go around the world, Europe, certainly the folks in AP as well are talking about sovereignty. They want to have control of their own digital future, particularly their AI future, their cloud future. Why is that? What are the drivers behind that?>> So I have to define digital sovereignty before I answer the why.
Dave Vellante
>> Yeah, great.>> It started off as data sovereignty. Where is the data? Data location, data in rest, data in use, data in motion. Great. But it very quickly shifted from that to operational sovereignty. Who runs my platform? Where is my control playing? Where are the keys and secrets? That had shifted towards technology sovereignty. Is the technology really open? And from there, much more recently in the last couple of years, it's been all about AI sovereignty. Where do my models run? Is my inferencing governed? Who has access at all times? So it kind of has gone a complete circle of digital sovereignty that includes data operational technology and AI. To answer your question on why, it comes down to two things, control and independence. But there's a why behind the control and independence. It's not like enterprises are trying to replicate being a technology vendor. They're not trying to replicate being a chip and a hardware vendor. That's not the point. They want operational resiliency. That is the fundamental reason why they want that control and independence is making sure their business is operational at all times. Their business is highly resilient from any type of media disruption external factors can bring in, and that's why they want the sovereignty. And in the world of AI, it is even more important because things are at a much, much accelerated pace. And with everything that generative AI and agents can do, it's very critical. Businesses don't lose that control, and that's why the control and independence is key. I would also say, sovereignty, there are different enterprises based on the region, how much of the shade of sovereignty they want. In some cases, it is pretty extreme where they want real control, real independence, where it's as far as I want it completely open source. I want to be able to run it. It needs to be in the region. And then even in some places it's probably on a lighter shade, but I want an open stack. I want a hybrid stack. I want it to be pluggable. But it all comes down to operational resiliency and business resiliency with AI operations.
Dave Vellante
>> And I want to come back to that openness, but your answer is very interesting to me because I think this is misunderstood. I think a lot of people equate sovereignty with GDPR and government regulation. Now maybe that was an initial catalyst. I mean, certainly the dominance of US cloud companies was sort of a catalyst as well, but it's evolved to much more than that, hasn't it?>> 100%. Regulations exist and it's critical to have the compliance around it. And companies do a lot of that using policies, and controls, and contracts. They do all of that, and that's critical. And every local regions and countries, there's different regulations that they do try to follow. To answer your question, it's expanded way beyond that. And yes, regulations address some of those, but everybody wants to make sure they are truly governed and the AI aspect of it is truly governed. Let me give you a simple example. Building agents is not rocket science. People can build agents, but-
Dave Vellante
>> I could build agents. And I'm not a rocket scientist.>> But the proliferation of agents is real in enterprises. Everybody is starting to build agents. And if you look at an organization, different departments, different organizations within the organization itself are starting to build agents. How do you as a CIO, how do you as a CTO make sure that there's a level of standardization, there is a level of consistency, there is a level of governance, there's a level of orchestration? All of these are internal fears, not necessarily that is coming just based on regulations. So yes, regulations is a factor, but companies do want to be able to make sure that they're running it in a highly governed manner, and that's also driving towards the need towards having the sovereignty.
Dave Vellante
>> When the public cloud really first started its ascendancy and organizations became comfortable with it, they quickly realized that the hyperscalers were actually better at security, for example, than most organizations. So they sort of leaned in, and we know the history there. But it seems to me that being able to control your operational model is a more complex piece. The cloud vendors aren't going to do that for you. They'll take a piece of it, but that means that organizations need solutions. So you guys have created the Sovereign Core you call it. Take us through the news. You have hard news. What is Sovereign Core? What is it all about?>> We officially made it generally available on Monday. Sovereign Core. It is a software platform. It is a software stack. And we were very cautious about making sure that it is a software solution and not a complete, "Here it is. This is what we're going to run." The reason is the second part in the four pillar sovereignty, the operational sovereignty. We want to give the flexibility for the enterprises to make a decision who they really want to be running that software. And it could be a vendor, if you choose a remote vendor that is sitting elsewhere, or it could be a local managed service provider. And we are working with multiple countries, local managed service providers, and we are equipping them to make sure that they are able to run the software in that region for the enterprise. The second part of it is we're also working with different ecosystem partners. We're working with Dell, AMD, Intel, and you'll see them through the day where we do want to give a turnkey solution to our clients. So those who want to get the software and bring their own hardware, and they've invested. Clients have been investing into their data center and they have their hardware investments. We just don't want to say, "No, start fresh. Don't leverage your existing investments. Expand that. Scale your AI using our existing investments using our Sovereign Core software and leverage your managed service providers to be able to, that we are partnering with to be able to run it locally there." That's your option. For those who are willing to make these hardware decisions, we're also giving you a turnkey option along with our ecosystem partners for you to be able to just plug and play and get started day one.
Dave Vellante
>> So this is a completely different IBM. I mean, historically, let's say go back 15, 20 years ago, IBM would say, "This is our hardware, our software." And kind of push it to customers. You've made a conscious decision not to do that. What went into that decision, obviously the openness, but take us through that decision and what did it require IBM to do both from a technical standpoint and a business model perspective?>> Simply put, it would be an anti-sovereignty pattern if we said, "Thou shall only run on our hardware, our software." So our principle was give the control and put the control back in the client's hands. To give that, we have to work with the ecosystem of partners. So that was a design philosophy from day one, and we built Sovereign Core from day one on one principle. We want enterprises to be able to scale AI quickly from their pilot to production. Today it takes 12, 18 months for a lot of enterprises, and a lot of projects fail. And we know why it fails. It's easier said than done. And we've had the learning over the last several years and we've seen what is actually keeping them from being able to scale. Our design philosophy was how do we get our clients productive in AI in a matter of hours rather than months and months of getting there. To do that, we have to make sure from day one everything all our decisions were cornered around one thing, give the control to the client, which means we have to work across the ecosystem with multiple partners and give them choices, flexibility, no vendor lock-in. The first principle we wrote down in our whiteboard when we were designing is no vendor lock-in. Be hardware-agnostic.
Dave Vellante
>> I got to push you on this. Okay. So you're not excluding IBM hardware, right?>> That's correct.
Dave Vellante
>> The folks in the infrastructure division, they can compete with, you mentioned Dell, Intel, AMD, et cetera. Does it run a little bit better on IBM? That's normally what would happen, right? You'd say, "Oh, well, we're open. You see this all the time in the industry. Wink, wink, but it runs a little bit better on IBM." Do you have little hooks in there that put their thumb on the scale for IBM? I mean, come on. What are the tricks you're pointing here?>> I love this question. So we do run on Fusion HCI as well, which is our hardware. And we do work with that, but we work on our hardware just as good as we work on any other hardware. So as a client, you make your decision, you make your choices, we will support them.
Dave Vellante
>> So let's get into a little deeper around Sovereign Core. What is it? It's a software only offering. What is it exactly?>> You're right. It is a software only offering, but we do provide turnkey solutions with our own hardware and with other ecosystem partners. Now what it provides you is three key pieces. One, a control plane. This control plane does everything from your secrets, your keys, your identity, your policy management, your access, all of that for an IT provider for you to stand up quickly. Then it provides a core set of base services. I do want to approach in a cluster, you want to provision an inferencing server, you want to provision a VM, it provides all of those capability. Then it has a catalog of services. And the catalog of services, we've been, again, being very cautious to make sure that we're giving clients a lot of options. So it could be IBM technologies, it could be open source technologies, or it could be third-party technologies. You can choose what you want and you are able to leverage and pull it out of this extensible catalog for you to build it. And where you would build it is in your tenant plane. So there are two key personas we're tackling. One is an IT service provider who sets up the control plane, brings up all of these base services, and then you turn it over to a line of business. And for the line of business, you could have multiple tenants in which you could be running your own agents, AI applications, and you're able to either pull it from the base service or from the catalog based on what the IT provider has made available. Yeah.
Dave Vellante
>> I'm assuming it's a hybrid control plane. If I want to have a hybrid environment, which is 99.9% of the environments, you can accommodate that. How do the clouds fit in? What are you hearing from customers about the hybrid nature of their cloud estates? How are they handling that and how are you handling that?>> So one thing that is super clear is enterprises are in this whole hybrid cloud journey. And IBM was probably in the forefront of predicting the world would be hybrid, and that's exactly what we're seeing. And hybrid also has expanded more recently, like multiple clouds. They just don't choose one cloud. So there's on prem, there is air-gapped, completely on prem, but air-gapped, regulated, highly regulated. Then you have single cloud, multiple clouds. So that solution, you can deploy it where you choose to because it is software stack and it is built in a way that you can deploy it anywhere. Now there is a fundamental architecture principle that we were able to leverage for us to be able to get this done in the time that we were able to. We've been building hybrid solution for years. And we made a conscious decision several years back, since our product can be deployed anywhere and we work across all the hyperscalers, we have to have a platform that we build that makes it build once, deploy anywhere. That was our mantra several years back, and that we built a platform internally. It was not meant to be externalized at all. It was meant where it has common metering, common logging, common observability, common billing, all of the common services. So we are able to take the common platform and be able to deploy it. When in the last couple of years, this whole AI governance, AI compliance, all of this accelerator, this digital sovereignty accelerated, we're like, guess what? We already have this platform that we've been using it. Let's take that and externalize it and make it a truly hybrid cloud solution for clients. That actually accelerated us in bringing Sovereign Core to market faster the right way. And we do believe our solution is Sovereign 2.0. There's 1.0, which has fallen short on a lot of things. And we've taken a very conscious effort on 2.0, and that was because of this platform that we had.
Dave Vellante
>> So build once, deploy anywhere, that's obviously the mantra of containers generally and Kubernetes specifically, but of course it's like that tweet. You don't just deploy Kubernetes. You have to harden it, and that's what Red Hat does. And so presumably that's a key part of the technology.>> Correct. Our software set comes with the Red Hat components as well. We use Red Hat AI. So as part of Red Hat AI, we have OpenShift in it. We have Quay, ACM. These are all technologies that come from Red Hat. And then we have our platform services, and then our AI services and other technologies.
Dave Vellante
>> Technically, you have a full stack solution that takes a lot of work, deep integration, co-design across hardware and software, but you've got this spectrum that you're accommodating of openness. What kind of technical challenges does that present for IBM? And I think it's obvious the benefit for customers, no lock-in choice, et cetera. But how did you accommodate both and get a solution to market relatively quickly?>> So I said three things we had to make sure we're giving the clients a choice. One, IBM technology to any third-party. Still proprietary, but third-party technology. Three is open source. Now we're talking about open source tools, models, services, all of that. But all of that has to work in a sovereign boundary. While we control, we control, it's very easy to bring it on a sovereign boundary. But the minute you say, "Bring your own model." And we actually give the flexibility, our model as a service, and Sovereign Core actually lets you bring your own model. Now that brings in a lot more challenges of how do you make sure that model can run in a sovereign boundary. Same thing our catalog. We have an extensible catalog with a lot of services. They're all not our products. They're ecosystem of partners that we have in the catalog, making sure they have the boundary. They're able to still create the boundary around the systems once it's deployed. We have to really design it in a way that it can keep that boundary created. That's probably the hardest challenge we had to tackle. Designing the solution itself was not the challenge. Understanding digital sovereignty, because we spent a good amount of time like, "Hey, we need to have a really Sovereign 2.0 kind of point of view around this, that most solutions we believe in the market is falling short of it." So those were not our challenge. Our biggest challenge was what we don't control, how do we still control to make sure it is programmable.
Dave Vellante
>> Interesting. Because IBM can certainly dial up Anthropic or OpenAI, or whomever, whatever, pick your LLM vendor, but there's so many open source models, particularly the work that is coming out of China is just amazing. So presumably many people want to tap that. You have to at least architecturally accommodate that. You can't test every model out there. There's millions of them. So what's the secret sauce there, particularly around security and governance? What's the architectural element or elements that enable that?>> We defined a very clear framework. I'm going to use Apple as an example. Not everybody can put anything into an Apple Store. There is a level of structure that you have to follow for you to actually make it in for it to run properly on ecosystem. People don't actually quite understand that in the beginning. It makes a lot of sense. There's a framework of what comes in so that you can continue. We followed a very similar principle. As much as we are giving you the flexibility, yes, you can, but at the same time, we make the commitment of trust around it and the boundary around it, and we have to make sure that that framework was followed. So if you're putting something into a catalog, it has to follow a certain aspect before you publish it in. Now the ones that we publish by ourselves, we test it a particular way, and we know that is harnessed towards what we make as commitment around governance and trust and security, all of that. And we do give the option as well. So the logs that you collect, like let's say you have a same provider that you're already using. Because one of the very fundamental design principles, when we wrote up, before we started building this, what are all our design principle? And one of the design principle was let customers leverage their existing investment. Don't go tell them you need a new security solution, you need a new governance solution. If they have on, we should be able to integrate into it and let them leverage the existing investment. That brings a bit of challenge on it, but again, we at least have tried to keep the door clean enough that if something comes, it's your decision. You are choosing, the control is in your hands, you are choosing what to go out and what to come in. And most importantly, it's visible. It's observable. You can audit it. You can trace it back what went in and what came out.
Dave Vellante
>> So you hear the app store of sovereign AI, and you set the standards and the policies as long as you adhere to them and you're adhering to them yourselves. You can then make a promise of governance and security and trust. And then so what happens when you run into a situation where that compatibility or those standards aren't met? Who's responsible? Is that something that you look at the market and you say, "Okay, it's worth it for us to do that engineering work or maybe the customer has to do it or somebody in the ecosystem"? Explain how that all works with the white space that presumably is out there and even further expands your TAM.>> The answer will be a combination all of that, and we have a very, very healthy ecosystem of partners. Where we do, we're going to work with them. And we have this concept called golden patterns. Again, one of the design clubs always said is these are going to be our golden patterns. So we're going to say the most successful, the most fastest, the most secure, here's a pattern we give you. But you can always choose a non-golden pattern to run. That is your choice. That is your flexibility to do that. But the golden patterns is what we test, and it could be with ecosystem, it could be with our solution. It could be something that we completely stand by, but we still give you the flexibility to go into a non-golden pattern if you choose to.
Dave Vellante
>> How unique is this? What would you say are IBM's clear differentiators in the market?>> One, genuinely believe we thought through a Sovereign 2.0. And the first thing is on the control plane, putting the control plane in your boundary was the first thing we made a decision on. Because if you look at most of the 1.0 kind of solutions, the control plane still sits somewhere else. Somebody else is still running it for you. Your keys and secrets are still sitting elsewhere. Yes, your data is there. Sure, it's all there, but some vendors somewhere still has the keys to the kingdom. So our first decision was put the control plane along with the data plane, tenant plane, everything, into the boundaries so the client has the keys to that. That's number one. Second, we're a company that's been operating across all the countries for decades. So we had the sovereign concept even before sovereign became such a big word. We knew how to work across the world globally in a way that is actually makes sense. We understand the regulations of the country. We understand the laws, the local policies, the local controls in place. Looking at data very differently to make sure it's not through policies, we can always demonstrate it by spending months and months with audits and contracts and all of that. We're like, no, our differentiator needs to be in a way where we can show continuous compliance where a client can see at any point in time, are they compliant or not? Two, it should be automated. Anything that we can automate, we do live in 2026, we should try to automate it. Three, how do we make sure that there is evidence, there is proof? It's sovereignty with receipts is what we say. And how do we make sure that we are able to give that evidence and proof at the go, at the ready for the client? So that would be the second thing on the whole ongoing compliance and automated and readiness and proof. The third thing was we know AI takes time. We want to be able to accelerate it. We also know why AI takes time. We know it comes down to the governance aspect of it, the scale aspect of it. It's not lack of tools. There's enough tools. There's enough tools anybody can build AI. But the problem is you're not able to do it in a way that you're comfortable for your enterprise at the scale you need, at the governance you need. How do you start generating AI in hours? So our simple funder was generate AI in ours with automated risk and compliance in a sovereign boundary, in a trusted manner. That was our one-liner, and everything we built was just following that principle.
Dave Vellante
>> I'm going to manifest my Hillary Clinton when she was the first lady take off on... It takes a village, it takes an ecosystem. So paint a picture of your ecosystem. What does that look like? What comprises the ecosystem?>> I'll have to say I'm very cloudy. Even at our GA and our press release, we had a section on IBM and then we had a equal section on our ecosystem partners, and that's a principle. The scale is through ecosystem. We established that several years ago. We scaled through our ecosystem and we're committed to it. Now, what is that ecosystem? In a world of sovereign, it's everything that fits into an enterprise. So it's storage. It's hardware. It's security solutions. It's other AI solutions. It's AI solutions that are models. It's AI for inferencing. It's AI for governance. It's the entire AI solution. Then it's data services. As much as we have our data services, we want to give you options. You want to use Postgres? Today you have a Postgres solution out of the box. If that's your favorite data solution, go for it. Hey, we do have data technologies. We could have just packaged that and sent it again, anti-sovereign pattern. So we're giving you that. So the ecosystem is anything from data, to AI, to hardware, to storage, to security. That is that extensible catalog of services that we're working with. And we have this partner ecosystem pretty well established, and we're leveraging that to expand. And we're also doing two things. We're listening to customers. We're talking to customers to understand what they want and see how we can work with their preferred choice of vendors, partners to bring things in. Two, open source is a big part. We're also choosing the right open source components into it so that they're able to get that open source. For those who are in the right side of the spectrum of sovereignty, they're able to. And if you choose to put something into your catalog, we're giving you a framework on how you can publish it and you can bring it into the catalog as well.
Dave Vellante
>> I mean, as we said up top, I mean, sovereignty is not a brand new thing, but we're entering a new phase. So talk to me about the go to-market. It seems like at this phase, a lot of it is belly to belly direct sales, but I could see over time having a marketplace evolve. Where are you with the go to-market? How does the go to-market work and how do you see it evolving?>> So to start with your spot on, we're taking just a two-pronged approach right now. We are going through our managed service providers. We're talking to a lot of MSPs around the world and who are immensely excited because they have a software solution now they can run on the hardware investments that they've had all along. They're super excited. So that is a big part of our go to-market is going through the MSPs. The second is through our own, of course, our hardware infrastructure through that angle and having a software solution and going that way, similarly through our partners who can take it to market as a complete turnkey solution. They have their AI hardware aspect. Now you've got a software AI aspect and go. That's in second aspect. Third is our own existing car, like customers that we've been talking to, that face-to-face still is a big part of our go to-market because they are the ones who actually drove the need towards this. We were listening to them consistently and regularly we could see. And people think this is a Europe thing or an Asia-Pacific thing. Yes, the word sovereign is a lot more deeper there, but regulated industries exist here. They're struggling with the same things of compliance, AI governance, being able to get AI to production. So across the world, our existing customers, we've heard from them, and this is a new way of solving some of their problems.
Dave Vellante
>> That's a great point about regulated industries. You're right. We tend to think about this as an overseas thing. It's not. Is every sovereign deployment sort of a Snowflake in terms of the pricing, the configuration? Are there any elements that are sort of standard? Is there a consumption model for your software or how does the pricing work?>> Now I understand you go to market question. What we have is resource units and the resource units are essentially a type of token if you want to see. So it is a consumption model. You buy a set of... You can pre-decide what you want to use, but we also give you an option of here's a bucket of stuff and you can choose what you want to deploy out of that. You can deploy a data service, imprinting service, something from the catalog, you can choose to deploy it. So yes, it's a consumption-based model. The consumption-based model is based on some resource units and the resource units can be pre-declared upfront or you can, as you use, you can report on what you're consuming.
Dave Vellante
>> Okay. So you've announced GA of Sovereign Core and...>> Oh, sorry, your question pricing, just-
Dave Vellante
>> Yeah, yeah, please.>> It is not a Snowflake.
Dave Vellante
>> It's not Snowflake.>> It is not a Snowflake. Because we have standard services and those are golden patterns, and that's what we're pricing it for.
Dave Vellante
>> Yeah. So a customer can understand what he or she is getting into at least in a reasonable range, small, medium, large.>> Right.
Dave Vellante
>> Okay. So what do you want to be able to say, let's say a year from now, that you're not able to say today about Sovereign Core?>> Oh, totally not prepared for it. What would I want to say from me here? I would actually want my customers to say more than what I say. Today I am saying I want to sit here with a customer, customers and multiple customers saying, "Hey, we finally have control of our solution. We have a great vendor that we can trust that we partner, but at the end of the day we have the control, we have the independence, and guess what? I can be in production and AI in matter of days. It's no longer a 12 to 18 month project." If I can sit with a bunch of customers and be able to say that, I think we've achieved our goal.
Dave Vellante
>> Well, I would love to sit down with you, Priya, in Think 2027 with a third chair, with a customer, and go through that. So, congratulations. I think this is a super exciting offering that IBM has developed. The ecosystem play is amazing. So, congratulations on getting that done. I know how much hard work goes into that.>> Thank you.
Dave Vellante
>> Appreciate your time. All right. And thank you for watching. This is Dave Vellante for theCUBE. If you found this video, you may have found it on YouTube. Go to thecube.net. We've got all the IBM Think 2026 videos. Go to siliconangle.com. I've written a number of articles as has our team. Thanks for watching, and we'll see you next time on theCUBE.
>> Digital sovereignty and AI sovereignty is one of the hottest topics that we hear about on theCUBE as we go around the world. My name is Dave Vellante, and I'm with theCUBE Research and SiliconANGLE and theCUBE. And we're here at IBM Think 2026 in this special studio that IBM has built for us. I appreciate that, IBM. And I'm here with Priya Srinivasan, who's the general manager of IBM software. And we're going to talk about sovereignty. Priya, thanks so much for spending some time with me.>> Of course. Thank you.
Dave Vellante
>> So as I said in my open, everywhere I go around the world, Europe, certainly the folks in AP as well are talking about sovereignty. They want to have control of their own digital future, particularly their AI future, their cloud future. Why is that? What are the drivers behind that?>> So I have to define digital sovereignty before I answer the why.
Dave Vellante
>> Yeah, great.>> It started off as data sovereignty. Where is the data? Data location, data in rest, data in use, data in motion. Great. But it very quickly shifted from that to operational sovereignty. Who runs my platform? Where is my control playing? Where are the keys and secrets? That had shifted towards technology sovereignty. Is the technology really open? And from there, much more recently in the last couple of years, it's been all about AI sovereignty. Where do my models run? Is my inferencing governed? Who has access at all times? So it kind of has gone a complete circle of digital sovereignty that includes data operational technology and AI. To answer your question on why, it comes down to two things, control and independence. But there's a why behind the control and independence. It's not like enterprises are trying to replicate being a technology vendor. They're not trying to replicate being a chip and a hardware vendor. That's not the point. They want operational resiliency. That is the fundamental reason why they want that control and independence is making sure their business is operational at all times. Their business is highly resilient from any type of media disruption external factors can bring in, and that's why they want the sovereignty. And in the world of AI, it is even more important because things are at a much, much accelerated pace. And with everything that generative AI and agents can do, it's very critical. Businesses don't lose that control, and that's why the control and independence is key. I would also say, sovereignty, there are different enterprises based on the region, how much of the shade of sovereignty they want. In some cases, it is pretty extreme where they want real control, real independence, where it's as far as I want it completely open source. I want to be able to run it. It needs to be in the region. And then even in some places it's probably on a lighter shade, but I want an open stack. I want a hybrid stack. I want it to be pluggable. But it all comes down to operational resiliency and business resiliency with AI operations.
Dave Vellante
>> And I want to come back to that openness, but your answer is very interesting to me because I think this is misunderstood. I think a lot of people equate sovereignty with GDPR and government regulation. Now maybe that was an initial catalyst. I mean, certainly the dominance of US cloud companies was sort of a catalyst as well, but it's evolved to much more than that, hasn't it?>> 100%. Regulations exist and it's critical to have the compliance around it. And companies do a lot of that using policies, and controls, and contracts. They do all of that, and that's critical. And every local regions and countries, there's different regulations that they do try to follow. To answer your question, it's expanded way beyond that. And yes, regulations address some of those, but everybody wants to make sure they are truly governed and the AI aspect of it is truly governed. Let me give you a simple example. Building agents is not rocket science. People can build agents, but-
Dave Vellante
>> I could build agents. And I'm not a rocket scientist.>> But the proliferation of agents is real in enterprises. Everybody is starting to build agents. And if you look at an organization, different departments, different organizations within the organization itself are starting to build agents. How do you as a CIO, how do you as a CTO make sure that there's a level of standardization, there is a level of consistency, there is a level of governance, there's a level of orchestration? All of these are internal fears, not necessarily that is coming just based on regulations. So yes, regulations is a factor, but companies do want to be able to make sure that they're running it in a highly governed manner, and that's also driving towards the need towards having the sovereignty.
Dave Vellante
>> When the public cloud really first started its ascendancy and organizations became comfortable with it, they quickly realized that the hyperscalers were actually better at security, for example, than most organizations. So they sort of leaned in, and we know the history there. But it seems to me that being able to control your operational model is a more complex piece. The cloud vendors aren't going to do that for you. They'll take a piece of it, but that means that organizations need solutions. So you guys have created the Sovereign Core you call it. Take us through the news. You have hard news. What is Sovereign Core? What is it all about?>> We officially made it generally available on Monday. Sovereign Core. It is a software platform. It is a software stack. And we were very cautious about making sure that it is a software solution and not a complete, "Here it is. This is what we're going to run." The reason is the second part in the four pillar sovereignty, the operational sovereignty. We want to give the flexibility for the enterprises to make a decision who they really want to be running that software. And it could be a vendor, if you choose a remote vendor that is sitting elsewhere, or it could be a local managed service provider. And we are working with multiple countries, local managed service providers, and we are equipping them to make sure that they are able to run the software in that region for the enterprise. The second part of it is we're also working with different ecosystem partners. We're working with Dell, AMD, Intel, and you'll see them through the day where we do want to give a turnkey solution to our clients. So those who want to get the software and bring their own hardware, and they've invested. Clients have been investing into their data center and they have their hardware investments. We just don't want to say, "No, start fresh. Don't leverage your existing investments. Expand that. Scale your AI using our existing investments using our Sovereign Core software and leverage your managed service providers to be able to, that we are partnering with to be able to run it locally there." That's your option. For those who are willing to make these hardware decisions, we're also giving you a turnkey option along with our ecosystem partners for you to be able to just plug and play and get started day one.
Dave Vellante
>> So this is a completely different IBM. I mean, historically, let's say go back 15, 20 years ago, IBM would say, "This is our hardware, our software." And kind of push it to customers. You've made a conscious decision not to do that. What went into that decision, obviously the openness, but take us through that decision and what did it require IBM to do both from a technical standpoint and a business model perspective?>> Simply put, it would be an anti-sovereignty pattern if we said, "Thou shall only run on our hardware, our software." So our principle was give the control and put the control back in the client's hands. To give that, we have to work with the ecosystem of partners. So that was a design philosophy from day one, and we built Sovereign Core from day one on one principle. We want enterprises to be able to scale AI quickly from their pilot to production. Today it takes 12, 18 months for a lot of enterprises, and a lot of projects fail. And we know why it fails. It's easier said than done. And we've had the learning over the last several years and we've seen what is actually keeping them from being able to scale. Our design philosophy was how do we get our clients productive in AI in a matter of hours rather than months and months of getting there. To do that, we have to make sure from day one everything all our decisions were cornered around one thing, give the control to the client, which means we have to work across the ecosystem with multiple partners and give them choices, flexibility, no vendor lock-in. The first principle we wrote down in our whiteboard when we were designing is no vendor lock-in. Be hardware-agnostic.
Dave Vellante
>> I got to push you on this. Okay. So you're not excluding IBM hardware, right?>> That's correct.
Dave Vellante
>> The folks in the infrastructure division, they can compete with, you mentioned Dell, Intel, AMD, et cetera. Does it run a little bit better on IBM? That's normally what would happen, right? You'd say, "Oh, well, we're open. You see this all the time in the industry. Wink, wink, but it runs a little bit better on IBM." Do you have little hooks in there that put their thumb on the scale for IBM? I mean, come on. What are the tricks you're pointing here?>> I love this question. So we do run on Fusion HCI as well, which is our hardware. And we do work with that, but we work on our hardware just as good as we work on any other hardware. So as a client, you make your decision, you make your choices, we will support them.
Dave Vellante
>> So let's get into a little deeper around Sovereign Core. What is it? It's a software only offering. What is it exactly?>> You're right. It is a software only offering, but we do provide turnkey solutions with our own hardware and with other ecosystem partners. Now what it provides you is three key pieces. One, a control plane. This control plane does everything from your secrets, your keys, your identity, your policy management, your access, all of that for an IT provider for you to stand up quickly. Then it provides a core set of base services. I do want to approach in a cluster, you want to provision an inferencing server, you want to provision a VM, it provides all of those capability. Then it has a catalog of services. And the catalog of services, we've been, again, being very cautious to make sure that we're giving clients a lot of options. So it could be IBM technologies, it could be open source technologies, or it could be third-party technologies. You can choose what you want and you are able to leverage and pull it out of this extensible catalog for you to build it. And where you would build it is in your tenant plane. So there are two key personas we're tackling. One is an IT service provider who sets up the control plane, brings up all of these base services, and then you turn it over to a line of business. And for the line of business, you could have multiple tenants in which you could be running your own agents, AI applications, and you're able to either pull it from the base service or from the catalog based on what the IT provider has made available. Yeah.
Dave Vellante
>> I'm assuming it's a hybrid control plane. If I want to have a hybrid environment, which is 99.9% of the environments, you can accommodate that. How do the clouds fit in? What are you hearing from customers about the hybrid nature of their cloud estates? How are they handling that and how are you handling that?>> So one thing that is super clear is enterprises are in this whole hybrid cloud journey. And IBM was probably in the forefront of predicting the world would be hybrid, and that's exactly what we're seeing. And hybrid also has expanded more recently, like multiple clouds. They just don't choose one cloud. So there's on prem, there is air-gapped, completely on prem, but air-gapped, regulated, highly regulated. Then you have single cloud, multiple clouds. So that solution, you can deploy it where you choose to because it is software stack and it is built in a way that you can deploy it anywhere. Now there is a fundamental architecture principle that we were able to leverage for us to be able to get this done in the time that we were able to. We've been building hybrid solution for years. And we made a conscious decision several years back, since our product can be deployed anywhere and we work across all the hyperscalers, we have to have a platform that we build that makes it build once, deploy anywhere. That was our mantra several years back, and that we built a platform internally. It was not meant to be externalized at all. It was meant where it has common metering, common logging, common observability, common billing, all of the common services. So we are able to take the common platform and be able to deploy it. When in the last couple of years, this whole AI governance, AI compliance, all of this accelerator, this digital sovereignty accelerated, we're like, guess what? We already have this platform that we've been using it. Let's take that and externalize it and make it a truly hybrid cloud solution for clients. That actually accelerated us in bringing Sovereign Core to market faster the right way. And we do believe our solution is Sovereign 2.0. There's 1.0, which has fallen short on a lot of things. And we've taken a very conscious effort on 2.0, and that was because of this platform that we had.
Dave Vellante
>> So build once, deploy anywhere, that's obviously the mantra of containers generally and Kubernetes specifically, but of course it's like that tweet. You don't just deploy Kubernetes. You have to harden it, and that's what Red Hat does. And so presumably that's a key part of the technology.>> Correct. Our software set comes with the Red Hat components as well. We use Red Hat AI. So as part of Red Hat AI, we have OpenShift in it. We have Quay, ACM. These are all technologies that come from Red Hat. And then we have our platform services, and then our AI services and other technologies.
Dave Vellante
>> Technically, you have a full stack solution that takes a lot of work, deep integration, co-design across hardware and software, but you've got this spectrum that you're accommodating of openness. What kind of technical challenges does that present for IBM? And I think it's obvious the benefit for customers, no lock-in choice, et cetera. But how did you accommodate both and get a solution to market relatively quickly?>> So I said three things we had to make sure we're giving the clients a choice. One, IBM technology to any third-party. Still proprietary, but third-party technology. Three is open source. Now we're talking about open source tools, models, services, all of that. But all of that has to work in a sovereign boundary. While we control, we control, it's very easy to bring it on a sovereign boundary. But the minute you say, "Bring your own model." And we actually give the flexibility, our model as a service, and Sovereign Core actually lets you bring your own model. Now that brings in a lot more challenges of how do you make sure that model can run in a sovereign boundary. Same thing our catalog. We have an extensible catalog with a lot of services. They're all not our products. They're ecosystem of partners that we have in the catalog, making sure they have the boundary. They're able to still create the boundary around the systems once it's deployed. We have to really design it in a way that it can keep that boundary created. That's probably the hardest challenge we had to tackle. Designing the solution itself was not the challenge. Understanding digital sovereignty, because we spent a good amount of time like, "Hey, we need to have a really Sovereign 2.0 kind of point of view around this, that most solutions we believe in the market is falling short of it." So those were not our challenge. Our biggest challenge was what we don't control, how do we still control to make sure it is programmable.
Dave Vellante
>> Interesting. Because IBM can certainly dial up Anthropic or OpenAI, or whomever, whatever, pick your LLM vendor, but there's so many open source models, particularly the work that is coming out of China is just amazing. So presumably many people want to tap that. You have to at least architecturally accommodate that. You can't test every model out there. There's millions of them. So what's the secret sauce there, particularly around security and governance? What's the architectural element or elements that enable that?>> We defined a very clear framework. I'm going to use Apple as an example. Not everybody can put anything into an Apple Store. There is a level of structure that you have to follow for you to actually make it in for it to run properly on ecosystem. People don't actually quite understand that in the beginning. It makes a lot of sense. There's a framework of what comes in so that you can continue. We followed a very similar principle. As much as we are giving you the flexibility, yes, you can, but at the same time, we make the commitment of trust around it and the boundary around it, and we have to make sure that that framework was followed. So if you're putting something into a catalog, it has to follow a certain aspect before you publish it in. Now the ones that we publish by ourselves, we test it a particular way, and we know that is harnessed towards what we make as commitment around governance and trust and security, all of that. And we do give the option as well. So the logs that you collect, like let's say you have a same provider that you're already using. Because one of the very fundamental design principles, when we wrote up, before we started building this, what are all our design principle? And one of the design principle was let customers leverage their existing investment. Don't go tell them you need a new security solution, you need a new governance solution. If they have on, we should be able to integrate into it and let them leverage the existing investment. That brings a bit of challenge on it, but again, we at least have tried to keep the door clean enough that if something comes, it's your decision. You are choosing, the control is in your hands, you are choosing what to go out and what to come in. And most importantly, it's visible. It's observable. You can audit it. You can trace it back what went in and what came out.
Dave Vellante
>> So you hear the app store of sovereign AI, and you set the standards and the policies as long as you adhere to them and you're adhering to them yourselves. You can then make a promise of governance and security and trust. And then so what happens when you run into a situation where that compatibility or those standards aren't met? Who's responsible? Is that something that you look at the market and you say, "Okay, it's worth it for us to do that engineering work or maybe the customer has to do it or somebody in the ecosystem"? Explain how that all works with the white space that presumably is out there and even further expands your TAM.>> The answer will be a combination all of that, and we have a very, very healthy ecosystem of partners. Where we do, we're going to work with them. And we have this concept called golden patterns. Again, one of the design clubs always said is these are going to be our golden patterns. So we're going to say the most successful, the most fastest, the most secure, here's a pattern we give you. But you can always choose a non-golden pattern to run. That is your choice. That is your flexibility to do that. But the golden patterns is what we test, and it could be with ecosystem, it could be with our solution. It could be something that we completely stand by, but we still give you the flexibility to go into a non-golden pattern if you choose to.
Dave Vellante
>> How unique is this? What would you say are IBM's clear differentiators in the market?>> One, genuinely believe we thought through a Sovereign 2.0. And the first thing is on the control plane, putting the control plane in your boundary was the first thing we made a decision on. Because if you look at most of the 1.0 kind of solutions, the control plane still sits somewhere else. Somebody else is still running it for you. Your keys and secrets are still sitting elsewhere. Yes, your data is there. Sure, it's all there, but some vendors somewhere still has the keys to the kingdom. So our first decision was put the control plane along with the data plane, tenant plane, everything, into the boundaries so the client has the keys to that. That's number one. Second, we're a company that's been operating across all the countries for decades. So we had the sovereign concept even before sovereign became such a big word. We knew how to work across the world globally in a way that is actually makes sense. We understand the regulations of the country. We understand the laws, the local policies, the local controls in place. Looking at data very differently to make sure it's not through policies, we can always demonstrate it by spending months and months with audits and contracts and all of that. We're like, no, our differentiator needs to be in a way where we can show continuous compliance where a client can see at any point in time, are they compliant or not? Two, it should be automated. Anything that we can automate, we do live in 2026, we should try to automate it. Three, how do we make sure that there is evidence, there is proof? It's sovereignty with receipts is what we say. And how do we make sure that we are able to give that evidence and proof at the go, at the ready for the client? So that would be the second thing on the whole ongoing compliance and automated and readiness and proof. The third thing was we know AI takes time. We want to be able to accelerate it. We also know why AI takes time. We know it comes down to the governance aspect of it, the scale aspect of it. It's not lack of tools. There's enough tools. There's enough tools anybody can build AI. But the problem is you're not able to do it in a way that you're comfortable for your enterprise at the scale you need, at the governance you need. How do you start generating AI in hours? So our simple funder was generate AI in ours with automated risk and compliance in a sovereign boundary, in a trusted manner. That was our one-liner, and everything we built was just following that principle.
Dave Vellante
>> I'm going to manifest my Hillary Clinton when she was the first lady take off on... It takes a village, it takes an ecosystem. So paint a picture of your ecosystem. What does that look like? What comprises the ecosystem?>> I'll have to say I'm very cloudy. Even at our GA and our press release, we had a section on IBM and then we had a equal section on our ecosystem partners, and that's a principle. The scale is through ecosystem. We established that several years ago. We scaled through our ecosystem and we're committed to it. Now, what is that ecosystem? In a world of sovereign, it's everything that fits into an enterprise. So it's storage. It's hardware. It's security solutions. It's other AI solutions. It's AI solutions that are models. It's AI for inferencing. It's AI for governance. It's the entire AI solution. Then it's data services. As much as we have our data services, we want to give you options. You want to use Postgres? Today you have a Postgres solution out of the box. If that's your favorite data solution, go for it. Hey, we do have data technologies. We could have just packaged that and sent it again, anti-sovereign pattern. So we're giving you that. So the ecosystem is anything from data, to AI, to hardware, to storage, to security. That is that extensible catalog of services that we're working with. And we have this partner ecosystem pretty well established, and we're leveraging that to expand. And we're also doing two things. We're listening to customers. We're talking to customers to understand what they want and see how we can work with their preferred choice of vendors, partners to bring things in. Two, open source is a big part. We're also choosing the right open source components into it so that they're able to get that open source. For those who are in the right side of the spectrum of sovereignty, they're able to. And if you choose to put something into your catalog, we're giving you a framework on how you can publish it and you can bring it into the catalog as well.
Dave Vellante
>> I mean, as we said up top, I mean, sovereignty is not a brand new thing, but we're entering a new phase. So talk to me about the go to-market. It seems like at this phase, a lot of it is belly to belly direct sales, but I could see over time having a marketplace evolve. Where are you with the go to-market? How does the go to-market work and how do you see it evolving?>> So to start with your spot on, we're taking just a two-pronged approach right now. We are going through our managed service providers. We're talking to a lot of MSPs around the world and who are immensely excited because they have a software solution now they can run on the hardware investments that they've had all along. They're super excited. So that is a big part of our go to-market is going through the MSPs. The second is through our own, of course, our hardware infrastructure through that angle and having a software solution and going that way, similarly through our partners who can take it to market as a complete turnkey solution. They have their AI hardware aspect. Now you've got a software AI aspect and go. That's in second aspect. Third is our own existing car, like customers that we've been talking to, that face-to-face still is a big part of our go to-market because they are the ones who actually drove the need towards this. We were listening to them consistently and regularly we could see. And people think this is a Europe thing or an Asia-Pacific thing. Yes, the word sovereign is a lot more deeper there, but regulated industries exist here. They're struggling with the same things of compliance, AI governance, being able to get AI to production. So across the world, our existing customers, we've heard from them, and this is a new way of solving some of their problems.
Dave Vellante
>> That's a great point about regulated industries. You're right. We tend to think about this as an overseas thing. It's not. Is every sovereign deployment sort of a Snowflake in terms of the pricing, the configuration? Are there any elements that are sort of standard? Is there a consumption model for your software or how does the pricing work?>> Now I understand you go to market question. What we have is resource units and the resource units are essentially a type of token if you want to see. So it is a consumption model. You buy a set of... You can pre-decide what you want to use, but we also give you an option of here's a bucket of stuff and you can choose what you want to deploy out of that. You can deploy a data service, imprinting service, something from the catalog, you can choose to deploy it. So yes, it's a consumption-based model. The consumption-based model is based on some resource units and the resource units can be pre-declared upfront or you can, as you use, you can report on what you're consuming.
Dave Vellante
>> Okay. So you've announced GA of Sovereign Core and...>> Oh, sorry, your question pricing, just-
Dave Vellante
>> Yeah, yeah, please.>> It is not a Snowflake.
Dave Vellante
>> It's not Snowflake.>> It is not a Snowflake. Because we have standard services and those are golden patterns, and that's what we're pricing it for.
Dave Vellante
>> Yeah. So a customer can understand what he or she is getting into at least in a reasonable range, small, medium, large.>> Right.
Dave Vellante
>> Okay. So what do you want to be able to say, let's say a year from now, that you're not able to say today about Sovereign Core?>> Oh, totally not prepared for it. What would I want to say from me here? I would actually want my customers to say more than what I say. Today I am saying I want to sit here with a customer, customers and multiple customers saying, "Hey, we finally have control of our solution. We have a great vendor that we can trust that we partner, but at the end of the day we have the control, we have the independence, and guess what? I can be in production and AI in matter of days. It's no longer a 12 to 18 month project." If I can sit with a bunch of customers and be able to say that, I think we've achieved our goal.
Dave Vellante
>> Well, I would love to sit down with you, Priya, in Think 2027 with a third chair, with a customer, and go through that. So, congratulations. I think this is a super exciting offering that IBM has developed. The ecosystem play is amazing. So, congratulations on getting that done. I know how much hard work goes into that.>> Thank you.
Dave Vellante
>> Appreciate your time. All right. And thank you for watching. This is Dave Vellante for theCUBE. If you found this video, you may have found it on YouTube. Go to thecube.net. We've got all the IBM Think 2026 videos. Go to siliconangle.com. I've written a number of articles as has our team. Thanks for watching, and we'll see you next time on theCUBE.
Dave Vellante
