By Loïc Decaux, Head of Research, Development & Innovation and Lead of the AI Lab.
Executive summary
Generative AI is no longer a futuristic promise. It is already embedded in our daily lives and increasingly present in organizations. Yet despite massive investments, growing experimentation, and technological capabilities, many organizations struggle to translate their AI ambitions into tangible business results.
Based on cross-sector insights shared during an Intys executive roundtable with AI and Transformation leaders that we hosted early December 2025, this white paper argues a simple but often overlooked truth: a Generative AI transformation is not primarily a technology challenge - it is an execution challenge.
Success depends far less on choosing the “right” model or platform than on the organization’s ability to rethink how work is done, how decisions are made, and how people adopt new ways of working. At the heart of this execution challenge lies a critical role: the AI Business Analyst — a role designed to bridge strategy, operations, technology, and people.
What leaders are really saying: Lessons from the field
During the Intys AI roundtable, leaders from sectors such as insurance, construction, defense, public services, end education shared similar experiences. Despite very different contexts, the same frustrations kept resurfacing.
One participant summarized it clearly:
“The real question is not what AI can do, but what it is for.”
“The real question is not what AI can do, but what it is for.”
Organizations are experimenting extensively. Proofs of concept, copilots, chatbots, and internal tools are multiplying. Yet only a limited number of these initiatives make it into daily operations or deliver measurable impact.
“We test a lot. We industrialize very little.”
“We test a lot. We industrialize very little.”
What emerges from these discussions is not a lack of technological maturity. On the contrary, most organizations already have access to powerful platforms and tools. The challenge lies elsewhere.
“Technology is rarely the bottleneck. Change management is.”
“Technology is rarely the bottleneck. Change management is.”
AI initiatives fail when they do not simplify work, accelerate decisions, or improve performance in a concrete and visible way. Too often, they remain disconnected from how people actually work.
Another recurring theme is the role of management — particularly middle management. Resistance is rarely explicit, but it is real. Fear of losing control, fear of being replaced, fear of having to abandon familiar tools and processes all slow down execution.
“The higher you go in the hierarchy, the more fears appear.”
“The higher you go in the hierarchy, the more fears appear.”
Add to this the complexity of governance, data quality, compliance, and security, and the result is clear: the difficulty is not imagining AI-powered futures, but executing them in real organizational environments.
The execution paradoxes of GenAI transformation
Why is execution so hard? Because GenAI transformations are shaped by tensions that cannot be resolved by technology alone.
Organizations are under strong pressure to move fast, yet they must remain compliant, secure, and trustworthy. Move too quickly, and fragmentation and shadow AI appear. Move too cautiously, and momentum is lost. These tensions require arbitration at the top — not technical fixes.
At the same time, AI promises innovation, but most organizations are built on legacy processes and systems. You cannot simply “add” AI on top of existing ways of working and expect transformation to happen.
“You don’t scale AI by adding tools. You scale it by redesigning how work is done.”
“You don’t scale AI by adding tools. You scale it by redesigning how work is done.”
There is also a constant tension between bottom-up energy and top-down alignment. Operational teams often see immediate opportunities for AI to help them. Executives seek coherence, prioritization, and risk control. Without clear arbitration, initiatives multiply, priorities blur, and value remains trapped.
Finally, there is the human dimension. Although AI is largely an augmentation tool, fear narratives dominate.
“We need a reassuring story: AI as an enabler, not a replacement.”
“We need a reassuring story: AI as an enabler, not a replacement.”
Without a clear and credible narrative about the role of humans in an AI-enabled organization, adoption stalls.
Why the AI Business Analyst is critical for execution
If GenAI transformation is an execution challenge, then it requires execution roles — not only technical ones. This is where the AI Business Analyst becomes essential.
The AI Business Analyst does not start with algorithms or platforms. The starting point is work itself: how tasks are performed, where time is lost, where decisions are slowed, where frustration accumulates.
“Capture the need, challenge how people work, and sit with them.”
“Capture the need, challenge how people work, and sit with them.”
This role is about asking the right questions before proposing solutions. It connects strategic ambition with operational reality.
AI strategies often remain abstract: vision statements, roadmaps, guiding principles. The AI Business Analyst translates these intentions into concrete, executable initiatives by identifying relevant use cases, clarifying expected value, and ensuring that solutions fit real workflows.
Just as importantly, this role helps navigate governance without killing momentum. It balances speed and control, innovation and compliance, standardization and local relevance. The objective is not to block initiatives, but to make them scalable, responsible, and sustainable.
Finally, the AI Business Analyst plays a decisive role in anchoring AI in daily operations. AI delivers value only when it is embedded into routines, decisions, and processes — not when it remains a separate layer.
“AI succeeds when it simplifies work, accelerates decisions, and improves performance.”
“AI succeeds when it simplifies work, accelerates decisions, and improves performance.”
This is not a technical integration problem. It is an execution problem.
Execution is at the core of Intys’ DNA and will always be so
At Intys, execution is not an afterthought. It is the foundation of how we approach transformation.
- Inspire means clarifying ambition and aligning AI initiatives with strategic priorities before discussing tools.
- Innovate means challenging processes and ways of working, not just deploying new technologies.
- Integrate means orchestrating data, governance, adoption, and delivery so that AI becomes part of everyday operations.
- Impact means measuring value and anchoring it in reality — not in presentations.
This is why Intys places the AI Business Analyst at the center of Generative AI transformations. Because execution is where strategy becomes real.
Final thoughts: Opening the debate
Generative AI is reshaping organizations faster than most operating models can adapt. The key questions leaders must now ask are not technical ones:
- Is our AI ambition truly executable?
- Do we understand how work needs to change — not just which tools to deploy?
- Have we invested enough in execution capabilities, not only platforms?
- Who is accountable for turning AI into real business value?
This white paper is not meant to provide all the answers. Its purpose is to reframe the conversation - from AI as a technology topic to AI as a strategy execution discipline.
If you would like to continue this debate, explore these questions further, or confront them with your own organizational reality, we would be pleased to exchange with you.
Because the future of Generative AI will not be decided by technology alone - it will be decided by how well we execute.
Loïc Decaux Head of Research, Development & Innovation Lead of AI Lab
Luis Parisot Partner Intys