When a homeowner in Budapest files an insurance claim, artificial intelligence handles everything. The AI system analyzes photos, reads documents, calculates losses and authorizes payments without human involvement. But the real challenge isn’t technical – it’s managerial.
Business schools across Europe are moving beyond basic AI literacy to teach executives something more complex: how to collaborate with intelligent machines. The focus has shifted from understanding what AI can do to knowing when humans should step in, when to trust the system, and who takes responsibility when things go wrong.
This shift matters because AI is no longer just a tool – it’s becoming a decision-making partner. Research published this year by MIT Sloan Management Review shows that humans and AI working together outperform either working alone. But that collaboration requires new management skills that traditional business education never had to teach.
At Uniqa Insurance Hungary, the NiQA system can settle claims within 24 hours. But chief executive Krisztián Kurtisz, who developed the project with Corvinus School for Executive Education, says the bigger change is organizational. “The real significance is the management model it demands,” explains László Eszes, the school’s director. Companies must rethink “the division of decision rights between human experts and AI systems.”
That division of labor is becoming the core curriculum at top business schools. At Insead near Paris, executives work through AI-enabled simulations where they must combine their judgment with machine-generated data. The goal isn’t to replace human decision-making but to sharpen it.
“Different ways of integrating AI lead to different outcomes, and leaders must understand when each is appropriate,” says Sameer Hasija, Insead’s dean of executive education. Participants often leave with ambitious plans, including one African company now developing a complete business model overhaul. The shift is from asking how to do things better to whether they should be done differently at all.
At HEC Paris, the focus is practical. The school’s AItelier platform walks executives through translating business problems into AI applications. One CFO wanted to build an AI assistant for real-time financial risk monitoring. What seemed like a technical project quickly became a series of management decisions:
- Who owns the data?
- How is data quality assured?
- Who is accountable for the system’s outputs?
“These agents don’t merely provide answers. They prompt reflection, challenge assumptions,” says David Restrepo, who heads AI executive education at HEC. The process ends not with automated solutions but with “collective refinement” – executives testing ideas with faculty and peers.
The need for human oversight becomes clearer once AI systems are deployed. At Essec Business School, programs focus on what happens after implementation. Models degrade as conditions change. Performance must be constantly reassessed. Decisions about retraining systems require balancing technical and commercial factors.
“Deploying an AI system is not the end of the story,” says Thomas Huber, academic director of Essec’s Executive Master in Artificial Intelligence program. The goal is developing “AI-fluent” leaders who can translate between technical and business teams and, when necessary, “push back” against AI recommendations.
This pushback capacity matters more than executives might realize. Recent MIT research warns that generative AI can act as a “persuasion engine,” subtly influencing how users interpret information. The challenge isn’t just whether AI is accurate, but whether its outputs are shaping judgment in ways executives don’t fully understand.
UPF Barcelona School of Management makes the stakes concrete. Workshops simulate AI systems making consequential decisions – rejecting job candidates, denying loans, reallocating medical resources. Participants must sign off on each outcome.
“What AI systems cannot do is sign. But someone must,” says executive education professor Giulio Toscani. When faced with that responsibility, executives quickly reassess their relationship with AI. Understanding how the system reached its conclusion becomes non-negotiable when a human is ultimately accountable.
The teaching methods themselves are evolving too. AI is now embedded directly into executive education programs. Insead uses tools like Lexarius for real-time feedback. HEC’s AItelier platform uses AI agents to challenge assumptions. Polimi Graduate School of Management in Milan deploys AI-powered avatars in virtual environments.
This evolution reflects a broader shift in how businesses think about AI. The early focus was on automation – replacing human tasks with machine efficiency. Now the emphasis is on augmentation – combining human judgment with machine capability. That requires a different kind of leadership, one that business schools are still learning to teach.
The Budapest insurance example illustrates both the promise and the challenge. NiQA can process claims faster and more consistently than humans. But someone still had to decide what threshold the system can authorize on its own, what happens when it encounters edge cases, and who takes responsibility when customers are dissatisfied with AI decisions.
Those are management questions, not technical ones. And they’re becoming the most important skills executives need to master as AI moves from experimental technology to operational reality.
