The most sober insight from the current AI debate is also the most practical: It is not the tool that determines success in the first place, but the process. Anyone who truly wants to scale AI must prioritize people, roles, data, and governance over model comparison.
What was supposed to be a morning arrival in Oslo turned into a late evening due to a canceled connecting flight. As the kickoff to the INSEAD Alumni Forum in Oslo, titled “Leadership, Balance, and Disruption in a New Era,” the experience nevertheless drove home a key point: Disruption is not just a buzzword, but first and foremost an operational reality.
You can't control everything. Even though management literature sometimes makes it sound as if all you have to do is pull the right framework off the shelf at the right time. All the better if the effort is still worth it in the end.
Description: Visit to the INSEAD Alumni Forum
That is exactly why it’s worth looking at AI not merely as a matter of tools, but as an issue of leadership and change.
Many AI projects fail not because the model isn’t capable enough, but because the organization isn’t prepared. Processes are unclear. Data is scattered. Roles aren’t defined. Business units expect magic, IT expects requirements, and management expects results—and somewhere in between, a pilot project emerges with a slick presentation but unclear operational details.
Artificial intelligence doesn't automatically make a bad system smart. It just makes it go bad faster.
That is why successful AI implementation does not begin with the question of which tool to procure. It begins with the question of which process to improve. Only then do data quality, responsibilities, the roadmap, governance, and quality assurance come into play.
This also changes the role of humans. In many business situations, AI doesn't simply replace decisions. It prepares them, speeds them up, provides feedback, or makes suggestions. That is precisely why humans remain essential.
Not as a decorative “human in the loop” who just gives a friendly nod at the end. But as a bridge between business processes and technology.
Employees need to understand what AI can do, where its limitations lie, and when results need to be reviewed. Managers need to understand which processes are affected, what risks arise, and what responsibilities cannot be delegated to a tool.
This requires a leadership culture that fosters transparency, skill development, and a certain degree of tolerance for mistakes. Not everyone needs to become an AI specialist. However, those in key roles need a solid foundational understanding of how AI can be used in a meaningful, safe, and verifiable way within their own work environment.
Another interesting point to consider comes from the alumni and MBA communities: Today, “buy-in” often has two meanings.
First, organizational buy-in—without which AI will remain stuck in the pilot phase. Simply agreeing isn’t enough. There needs to be a genuine understanding of why a process is being changed, who is affected, and how responsibilities will be allocated in the future.
Second, entering the business world by acquiring an existing company rather than starting a new one. Here, too, it’s clear that those who acquire and scale existing companies are never starting from scratch. They work with real processes, established structures, existing data, existing teams, and—occasionally—surprisingly long-lived Excel files.
That’s exactly where it’s decided whether AI will succeed—not in the innovation lab, but on the job.
These observations point to a simple yet challenging agenda: AI is not primarily a procurement project, but a transformation project. It will be successful where leadership fosters process clarity, engagement, and a capacity for learning.
AI projects rarely fail solely because of the technology. They often fail because processes, roles, and responsibilities aren't clearly defined.
Anyone who wants to use AI effectively should therefore start not with the tool, but with the organization. First, clarity in processes; then, clarity in data; then, accountability; and finally, scaling.
K11 helps companies view AI not as an isolated technology project, but as part of modern governance: structured, practical, and compatible with day-to-day business operations.