When ChatGPT was launched in late 2022, much of the business world entered a phase of intense experimentation. Companies tested dozens of tools, created innovation labs, and multiplied proof of concepts. Three years later, one question now dominates discussions: how can these experiments be transformed into a real competitive advantage?
The Boston Consulting Group report, AI at Work: Why Strategy Matters More Than Tools, provides a clear answer. According to this study conducted among nearly 12,000 employees in fourteen countries, company performance no longer depends primarily on the level of sophistication of the tools used. The organizations achieving the best results are those based on a clear and coherent strategy.
This conclusion echoes the insights from years of research conducted as part of the EntrepreneurIA project. Through more than a hundred interviews with executives, founders, experts, and decision-makers, the same reality emerges: AI is no longer a technological issue. It is becoming a matter of governance, organization, and transformation of business models.
The Miracle Tool Trap
One trend recurs consistently in the testimonials collected for EntrepreneurIA.
Many organizations begin their transformation with technology when they should start with business objectives.
The enthusiasm generated by generative models has sometimes led certain companies to adopt solutions before they have clearly identified the problems to solve or the success indicators to track.
This situation recalls the early days of digital transformation. The companies that created lasting value were not necessarily those with the best software. They were those that had rethought their processes and ways of operating.
The BCG report confirms this observation. A clear strategy produces more impact than an accumulation of tools. Artificial intelligence is not a strategy. It is merely a means serving a broader ambition.
Adoption Progresses Faster Than Value Creation
One of the most striking insights from EntrepreneurIA research concerns the gap between adoption and value creation.
The vast majority of companies surveyed now use artificial intelligence tools in certain activities. The observed gains are real: reduction in time spent on certain tasks, improvement in documentary quality, acceleration of analyses, automation of administrative processing, or enhancement of customer relations. Nevertheless, one observation stands out.
To date, among the companies observed as part of EntrepreneurIA, none has provided fully documented evidence of an overall return on investment from artificial intelligence. The observed benefits remain primarily measured at the level of certain processes or teams. Benefits are visible individually or on specific processes. However, their translation into consolidated gains in revenue, profitability, growth, or company valuation remains difficult to measure.
This observation indirectly aligns with BCG’s conclusions. Adoption is progressing rapidly. Uses are multiplying. But the ability to transform these scattered gains into sustainable economic performance remains a major challenge.
The New Blind Spot: Measuring Created Value
Why is it so difficult to measure the ROI of AI?
The answer lies partly in the very nature of this technology.
It is relatively simple to evaluate the time saved by an employee thanks to a generative assistant. It is much more complex to establish the link between this time saved and the actual improvement in the company’s financial results.
AI often acts on indirect variables: decision quality, execution speed, customer satisfaction, innovation, error reduction, or improved collaboration.
These benefits are real, but their economic translation requires management tools that few organizations yet possess.
One of the paradoxes observed today is therefore this: companies know how to measure AI usage, but they still struggle to precisely measure the value it creates.
AI Reveals Organizational Quality
Throughout the interviews conducted for EntrepreneurIA, another trend has emerged.
Artificial intelligence acts as an organizational revealer. In companies where processes are clear, responsibilities well-defined, and objectives shared, AI projects progress quickly.
Conversely, when processes are unclear or fragmented, AI tends to amplify existing dysfunctions. This observation aligns with BCG’s conclusions.
Individual gains can be significant. But without parallel organizational evolution, these gains often remain isolated and difficult to transform into sustainable competitive advantage. AI does not compensate for organizational weaknesses. It makes them more visible.
Human Skills Gain Value
One of the most reassuring insights from this transformation period concerns the role of humans. Contrary to the most alarming scenarios, observations collected in EntrepreneurIA show that companies are seeking more human skills.
Repetitive tasks are gradually being automated. On the other hand, skills related to judgment, creativity, critical analysis, decision-making, and understanding contexts are becoming more important. Value is shifting.
For a long time, competitive advantage largely rested on the ability to produce and execute quickly.
Tomorrow, it will rest more on the ability to interpret, arbitrate, imagine, and give meaning.
The BCG report confirms this evolution by showing that employees already spend more time supervising and directing AI systems than executing certain tasks themselves.
AI Agents Change the Nature of the Debate
A new stage is opening with the emergence of AI agents.
The first generative assistants helped employees produce content or search for information. Agents go further. They execute tasks, coordinate processes, and interact with multiple systems.
This evolution profoundly changes the stakes. The question is no longer just how to use AI. It becomes: how to govern systems capable of acting?
Companies interviewed as part of EntrepreneurIA largely share the same concerns.
How to ensure traceability of decisions? How to ensure regulatory compliance? How to distribute responsibilities? How to maintain trust?
These questions appear as important today as technical performance itself.
The Real Challenge: Reinventing the Company
The main insight emerging from both the BCG report and EntrepreneurIA work is probably this: the real transformation is not technological.
It is organizational. The most advanced companies are no longer just seeking to automate the existing.
They are beginning to rethink their operating methods around the new capabilities offered by artificial intelligence. The difference is considerable. Automating a process generally saves time. Redesigning that process can transform a business model.
This distinction explains why some organizations achieve spectacular results while others remain stuck in a succession of pilot projects.
From Generative AI to the Augmented Enterprise
2023 was the year of discovery.
2024 was the year of experimentation.
2025 marked the progressive industrialization of uses.
2026 seems to inaugurate a new phase: that of the augmented enterprise.
In this new stage, performance will no longer depend primarily on the quality of tools used.
It will depend on leaders’ ability to articulate strategy, governance, skills, corporate culture, and technologies.
Observations made as part of EntrepreneurIA lead to a conclusion that aligns with BCG’s.
AI adoption is now widely underway. Individual gains are visible. Operational benefits are multiplying. But demonstrating an overall return on investment at the enterprise level remains largely to be built.
This is precisely where the next challenge for leaders lies. The issue is no longer using artificial intelligence. The issue is now proving, measuring, and managing the value it actually creates.
Companies that succeed in this stage will have a considerable competitive advantage in the coming years.
Pascale Caron
AI Strategy Advisor | International Keynote Speaker | Co-author of EntrepreneurIA
Conferences • Executive Advisory • AI Governance
