2021.AI

One of the biggest reasons AI projects fail isn’t the models — it’s the foundation. Complex infrastructure, unclear governance, and compliance risks often stall promising initiatives before they can create real value. That’s why I am so excited about our new partnership with Sentia. Together, we’ve launched a fully managed AI platform that removes these barriers and gives enterprises a secure, scalable, and compliant foundation for AI. By combining Sentia’s hybrid cloud expertise with 2021.AI's GRACE Platform, enterprises can: - Speed up time-to-production for AI projects - Apply AI safely even to sensitive data - Enable internal teams to focus on innovation instead of operations This is a great step forward for Sovereign AI — creating environments where enterprises can innovate with confidence, knowing their data and models remain secure, governed, and in a location they trust. Marius Jeppesen Kim Lars Jakobsen Casper Hansen Bjørn Olesen Ditte Stage Mikael M. #SovereignAI #ResponsibleAI #AIGovernance #HybridCloud #AIAdoption #Partnership

I'm thrilled to have provided a small contribution to Ray Eitel-Porter, Dr Paul Dongha, and Miriam Vogel's new book Governing the Machine, now available. Thank you for mentioning 2021.AI and our GRACE AI Platform 📘 It's encouraging to see the market increasingly recognize that AI platforms must integrate both operations and governance. Too many solutions are "either/or": strong governance tooling without operational depth, or robust MLOps/LLMOps without embedded governance. That's suboptimal - you need both. AI is fundamentally different from technologies organizations have scaled before. You can't simply bolt on a governance layer; it must be embedded in model development, deployment, monitoring, and improvement - as part of the operating model, not an afterthought. A responsible approach to AI has never been more critical, which is why I deeply appreciate efforts like this book that bridge theory and execution. Thank you for the timely insights on AI Governance, and congratulations on an outstanding book, Ray, Paul, and Miriam! 👏 #aigovernance #responsibleAI #AI4good #GRC #AIGRC

AND what a nice and cool way to sense of complex health data 📊 🙌 How do we make complex health data easier accessible and available? That was the challenge set by the Danish Diabetes Association (Diabetesforeningen) to 2021.AI? The solution: A conversational AI agent that turns validated data into simple insights. The project was a successful collaboration with #IQVIA Nordics and 2021.AI. The solution enables citizens, journalists, and decision-makers to ask questions about diabetes data—and receive clear, data-driven answers and visualizations in real time and in plain language 💬 Tanja Thybo, Head of Research & Knowledge at the Danish Diabetes Association, puts it: "By bringing AI into our work with health data, we can lower the barriers to insight and give patients, journalists, and policymakers new ways to understand—and act on—the data we already have." As an example, a citizen can ask: “How many people are living with type 2 diabetes in the Region of Southern Denmark—and how has that changed over the past five years?” The AI finds, analyzes, answers (in plain language), and provides you with a visualization to show trends - imminently 💡 Powered the GRACE AI Platform, the solution has AI governance embedded🛡️ all answers are within the validated domain and is monitored to ensure accuracy, transparency, and trust. Want to see it in action? Try it here (and explore current diabetes trends) 👉 https://diabetestal.nu/ (Available in Danish only) Or read the full client story 👉 https://lnkd.in/d9s3j8gw

🚨 A drop from 43% to 27% - of leaders who say they trust fully autonomous agents, this dramatic drop in just one year according to Capgemini Research Institute, Sep 2025 “Rise of agentic AI“ At Agentdagen in Copenhagen yesterday I shared a perspective on how to close this gap. It calls for clear structures and oversight, a few steps: ✅ An AI register to eliminate shadow AI ✅ Full lifecycle management of all AI initiatives ✅ Cross-functional collaboration – in one unified platform ✅ Implementation of controls based on industry standards + internal AI policies Another perspective shared was, how that we at 2021.AI use Agentic AI ourselves – to keep processes transparent and under control. With the right solutions you can ensure that, it’s not oversight that slows you down. It’s the absence of it. The organizations that succeed with agentic AI will be the ones that embed structure and governance from the outset. Read more on👉 www.2021.ai #agenticai #aigovernance

🚨 𝗪𝗲’𝘃𝗲 𝗦𝗲𝗲𝗻 𝗧𝗵𝗶𝘀 𝗕𝗲𝗳𝗼𝗿𝗲. 𝗪𝗲 𝗜𝗴𝗻𝗼𝗿𝗲𝗱 𝗜𝘁 𝗧𝗵𝗲𝗻. 𝗪𝗶𝗹𝗹 𝗪𝗲 𝗜𝗴𝗻𝗼𝗿𝗲 𝗜𝘁 𝗔𝗴𝗮𝗶𝗻? Some decades ago, companies rushed to adopt best-of-breed ERP systems—each one a specialist tool, often from different vendors, tailored to solve a specific business need. 𝗪𝗵𝗮𝘁 𝗳𝗼𝗹𝗹𝗼𝘄𝗲𝗱 𝘄𝗮𝘀 𝗶𝗻𝘁𝗲𝗴𝗿𝗮𝘁𝗶𝗼𝗻 𝗵𝗲𝗹𝗹. Disconnected systems. Data silos. Fragmented workflows. Shadow IT. A lack of unified oversight. And above all: a massive governance gap. Few organizations had the structures in place to manage the complexity—and the cost of trying to regain control was steep. Fast forward to today: 𝗪𝗲'𝗿𝗲 𝗻𝗼𝘄 𝘄𝗶𝘁𝗻𝗲𝘀𝘀𝗶𝗻𝗴 𝗮 𝘀𝗶𝗺𝗶𝗹𝗮𝗿 𝗽𝗮𝘁𝘁𝗲𝗿𝗻 𝗲𝗺𝗲𝗿𝗴𝗲 𝘄𝗶𝘁𝗵 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜 𝗮𝗻𝗱 𝗮𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀 𝗮𝗴𝗲𝗻𝘁𝘀. Organizations are enthusiastically deploying AI models and agents across departments and functions. One for content. One for customer support. Another for analysis. Yet another to summarize meetings. Each doing their own thing. Each evolving in its own way. 𝗕𝘂𝘁 𝘄𝗵𝗲𝗿𝗲’𝘀 𝘁𝗵𝗲 𝗴𝗼𝘃𝗲𝗿𝗻𝗮𝗻𝗰𝗲? • Where are the controls? • The audit trails? • The coordination between autonomous systems? • The accountability if something goes wrong? Just like with ERP, the signs are already there: • Agents acting unpredictably • AI decisions that no one can fully explain • Overlapping tools with no ownership • Business leaders unsure who’s responsible for what And yet—very few are seriously addressing the governance layer. If this gap remains unaddressed, the costs will be immense. So will the risks. We’ve been here before. 𝗪𝗲 𝗸𝗻𝗼𝘄 𝗵𝗼𝘄 𝘁𝗵𝗶𝘀 𝘀𝘁𝗼𝗿𝘆 𝗲𝗻𝗱𝘀 𝘄𝗵𝗲𝗻 𝗰𝗼𝗻𝘁𝗿𝗼𝗹 𝗶𝘀 𝗮𝗻 𝗮𝗳𝘁𝗲𝗿𝘁𝗵𝗼𝘂𝗴𝗵𝘁. The question is: will we make the same mistake again? #AI #Governance #GenAI #EnterpriseTechnology #DigitalRisk #TechLeadership #LessonsFromThePast

AI: Implement or Employ? NVIDIA CEO Jensen Huang recently said IT will become the HR for AI agents. I agree with this analogy, but in my experience, it goes far beyond IT. LLM-based chat-agents are already becoming part of the workforce, engaging with customers and making decisions on behalf of companies. I discussed this with Peter Sondergaard and it raises the question: 💡 If AI agents are part of the workforce, what does it take to manage them? Success with AI requires us to look beyond implementation—We have to look at AI as if we employ it, defining its role, responsibilities, mandate and ensuring the AI agents are aligned with our company strategy. Just like human employees, AI agents need leadership. ✔️ Who sets their boundaries and measures performance? ✔️ How do we ensure they operate responsibly? ✔️ What new leadership tools are required in an AI-driven workforce? The shift is happening. Are we ready to lead AI, not just deploy it? Watch the full AI Watch here https://lnkd.in/g4StUzCm #AI #AILeadership #GenAI #AIGovernance