AI in Leadership: From Costly Experiment to Strategic Imperative

This article is based on Episode 32 of CfL's podcast, The Leadership Room, featuring Anders Hvid from Dare Disrupt. In the episode, he shares his experiences of implementing AI in Danish organizations and discusses how leaders can turn AI from an experimental technology into a source of real business value. Listen to the full conversation here.

By Christina Nielsen, june 2026

Artificial Intelligence Isn't New, But the Rules Have Changed

Artificial intelligence has been around as a concept since 1956, when Marvin Minsky—widely regarded as one of the founding figures of AI—first introduced the term at a scientific conference. Since then, AI has quietly become part of everyday life, powering everything from spam filters and navigation systems to Netflix recommendations. As Minsky famously put it: "As soon as it works, no one calls it AI."

What has changed, however, is the technology itself. The emergence of large language models such as GPT has fundamentally reshaped the role of AI in leadership and business. Where AI projects once required significant investment, specialist expertise, and large volumes of structured data, today's generative AI is widely accessible, affordable, and in many cases, free to use.

"AI has gone from being niche and expensive to broad and affordable," says Anders Hvid, founder of Dare Disrupt, who has spent more than a decade helping organizations implement emerging technologies and has delivered over 300 keynote presentations on the subject worldwide.

For today's leaders, this raises an urgent question: How do you approach a technology with the potential to transform almost every type of knowledge work within your organization?

Proven Impact: Leaders Can't Afford to Wait

The potential is no longer theoretical. Anders Hvid points to a landmark study conducted by Harvard Business School in collaboration with Boston Consulting Group (BCG), in which 750 consultants were divided into three groups and asked to solve the same business case. Those with access to AI delivered significantly better results:

• 12.5% more tasks completed

• 25% faster task completion

• 40% higher-quality outputs

"This represents a significant productivity gain. As a responsible leader, you simply can't afford to ignore it," Hvid emphasizes. He continues:

"If you're a leader and you don't invest time in understanding what this technology is and how to unlock its potential, I would consider that irresponsible."

The message is clear: AI in leadership is no longer a question of whether to adopt it, but how to use it effectively.

The article continues below.

"We've become accustomed to calculators, where two plus two always equals four. But with this type of IT system, we're dealing with probability rather than certainty—which means two plus two won't always equal four," Hvid explains.

This creates two key challenges that leaders need to navigate:

• Bias: Large language models can reproduce biases present in their training data. For example, an AI might automatically assume that a doctor is male while a nurse is female, reflecting stereotypes rather than objective reality.
• Hallucinations: Large language models can sometimes generate information that sounds convincing but is factually incorrect. This is a natural consequence of how generative AI works, making critical thinking and human judgment essential when evaluating its output.

A New Paradigm Requires New Leadership Capabilities

The result is that critical thinking becomes a core capability for anyone using AI in their work. Hvid recommends that organizations establish new quality assurance processes, such as a "four-eyes principle," where all AI-generated content is reviewed by a second person before it is shared with customers or published externally.

Data privacy is another important consideration. "OpenAI is not GDPR-compliant," Hvid notes. As a result, confidential information and personal data should never be entered into public large language models.

Three Practical Steps to Get Started with AI in Your Organization

1. Select a Pilot Group: Don't roll out AI across the entire organization from day one. Instead, start with a small group of employees who can experiment in a controlled environment, allowing you to capture insights, refine your approach, and build organizational learning before scaling.
2. Learn Prompting—It's a New Core Skill: Most people start by using AI like a search engine, but that's not where its real value lies. Prompting—the ability to give AI clear, structured instructions—is a new skill that requires practice. The better your prompts, the more useful and relevant the output will be.

3. Establish Clear Guidelines from Day One: Define exactly what AI can and cannot be used for within your organization. Be explicit about data privacy, bias awareness, and quality assurance. Remember: your employees are likely already using AI tools—whether your organization has formal policies in place or not.

A Practical Example: Multilingual Customer Service

Hvid shares a practical example from a property management company serving tenants from a wide range of cultural and linguistic backgrounds. The company used ChatGPT to help staff draft clearer, more polite, and more professional responses across multiple languages—even when replying to hostile or confrontational messages. The result was faster response times, improved customer service, and reduced resource requirements.

The Strategic Question: What Are We Actually Trying to Achieve?

Perhaps Anders Hvid's most important point isn't about the technology itself—it's about the strategic conversation it should spark. If AI simply helps us work faster, but we fill the time we save with even more tasks, have we actually gained anything?

"Most of us never get through everything on our to-do list. So if we don't stop and think about it, we'll simply go from having 160 open tasks to 140—and we won't even notice that we've become more productive," Hvid says.

Instead, Hvid argues that the greatest opportunity lies in improving the quality of work, rather than simply increasing the speed at which it gets done. AI should be used to raise the standard of existing work—not just to produce more of it.

There is another dimension leaders cannot afford to overlook: employee concerns. Will AI replace my job? According to Hvid, this conversation should be addressed proactively from day one. Leaders need to be transparent about the purpose of AI, how it will be used, and the role employees will play as the technology is introduced.

"There is a great deal of uncertainty and concern among employees, and that's something leaders need to address. You might as well tackle it head-on," Hvid recommends.

The Next Step: From Public AI Tools to an Organization-Wide AI Strategy

Episode 32 explores the first stage of AI adoption in leadership: using public, general-purpose language models to support everyday work. But that's only the beginning. In the following episode—Episode 33—Anders Hvid and host Henrik Eriksen take the conversation a step further, exploring how organizations can develop and train their own AI solutions tailored to their specific needs, workflows, and data.

Ready to Take the Next Step with AI in Your Leadership?

We offer courses and learning programs that equip leaders to navigate the AI era—from building a strategic understanding of AI to implementing it effectively in practice. Explore, for example, our course: GenAI i strategisk ledelse or find the learning program that's right for your organization's needs.

Explore our full AI course portfolio here