Keepit: The foundation AI can't build itself — a single point of truth

Date and Time: October 20 at 2:20 p.m. (EDT) 
Room: Future Directions Stage
Session code: FD10

AI doesn't create truth. It amplifies whatever it's given. As organizations expand AI across their SaaS applications, one lesson keeps resurfacing: you can't build a recovery strategy on the systems you need to recover from. This session explores why an independent, verified, and always-recoverable single point of truth is the foundation responsible AI adoption depends on — and why trust, recoverability, and AI strategy belong in the same conversation. 

What you’ll take away:

• A clearer understanding of why recovery responsibility has always sat with the organization — not the platform  
• Practical insight into the SaaS data protection gaps that AI adoption makes more urgent, not less  
• A framework for building a single point of truth that is independent, verified, and always recoverable  

Sal 5
Session code: [LSP06-P]

AI is on every other agenda slide today. Jakob Østergaard, CTO at Keepit, strips away the hype and shows what's genuinely new about AI risk, and what's simply the same questions that have always mattered. You'll leave with a better understanding of the AI technology foundation, letting you see through the noise and focus on what actually matters when turning opportunity into real improvements while keeping risk in check. 

Sovereignty first: The data foundation AI can't scale without

Date and Time: Oct 28, 2026 | 4:15–4:35 p.m.
Location: Gold stage

On Oct. 28 at 4:15 p.m. on the Gold Stage, Kim Larsen, group chief information security officer (CISO) at Keepit, presents "Sovereignty first: The data foundation AI can't scale without." He argues that AI readiness starts with data sovereignty and governance, not model selection, unpacking what it takes to know where your data lives, who can access it, and whether you can still reach it when a provider, region, or geopolitical situation changes. For partners advising customers on AI adoption, this session reframes governance as the prerequisite, not the afterthought.

Every AI strategy runs into the same wall: control over data. In this 20-minute session, Kim Larsen, Group CISO at Keepit, argues that AI-readiness starts with data sovereignty and governance, not model selection. He unpacks the practical difference between application sovereignty and data sovereignty (why the former is out of reach for most organizations in the short term, while the latter is not), and what it takes to know where your data lives, who can access it, and whether you can still reach it when a provider, region, or geopolitical situation changes. For partners advising customers on AI adoption, this session reframes governance as the prerequisite, not the afterthought.

Keepit: The amplified case for data sovereignty in the age of AI

Date and Time: November 11 at 11:30 a.m. (CET)
Room: 117 
Session code: SPS53

Data integrity is the precondition for trustworthy AI. When the data feeding your AI is incomplete or detached from its origin, the outputs carry that uncertainty. That integrity depends on sovereignty — the ability to access, recover, and verify your data independently. Leave this session with a practical framework for evaluating true recoverability, and a stronger sense of why sovereignty, governance, and AI strategy are inseparable. 

You will leave this session with: 

• A clearer understanding of why data sovereignty is foundational to a trustworthy single point of truth 
• A framework for evaluating whether your organization can truly access and recover its data independently of the platforms it runs on 
• Practical insight into how SaaS environments quietly erode data control — and what the accountability gap looks like in practice 

Room: Maritime Suite 25 
Session code: SPS1 

As organizations depend more on SaaS and AI, data sovereignty is no longer just about where data resides, but who controls it. I&O leaders will learn how to assess their ability to access, verify, and recover critical data independently. Drawing on real-world resilience experience, this session explores how independent recoverability can reduce cloud dependency, strengthen operational resilience, and support an AI-driven future. 

You will leave this session with: 

• Why data sovereignty increasingly means control and recoverability, not just residency.  
• How SaaS and AI dependencies are changing the operational risk landscape.  
• How to test whether your organization can independently access and recover critical data. 

Room: Theatre 3 
Session code: TH30  

As AI increases our reliance on SaaS data, what happens when that data is deleted, compromised, or unavailable? This session walks through practical SaaS recovery scenarios and the operational decisions behind them. See how independent access, verified recovery, and a trusted copy of critical data can help I&O teams protect the data AI depends on, reduce recovery risk, and strengthen operational resilience. 

Takeaways 

• See three SaaS recovery scenarios where AI and growing SaaS dependency challenge common recovery assumptions. 
• Understand what independent recovery requires when critical data is deleted, compromised, or unavailable. 
• Learn how to test recoverability and identify gaps before they impact SaaS operations and AI initiatives. 

When the cloud goes dark — resilience lessons from the field

Date and Time: November 18 at 10:55–11:15
Location: Blue Stage (Blå Scene) 

Hybrid threats don't announce themselves. They combine cyberattacks, infrastructure sabotage, and geopolitical pressure — and cloud services are a primary target. In this session, Morten Aagaard walks through real-world scenarios where SaaS availability fails and what that exposes for organizations built on hyperscaler infrastructure. 

The session makes the case for keeping backup data outside the platforms it protects, in a location where data residency meets the regulatory requirements that apply to your organization. For European organizations, that means solutions built for European sovereignty requirements — not adapted from global hyperscaler stacks. 

Attendees will leave with a clearer picture of where single-vendor dependencies create blind spots, why data sovereignty is a resilience issue as much as a compliance one, and how independent cloud backup changes the recovery equation when primary access is lost. 
 

The session will be held in Danish.