A Systemic Technological Revolution
Generative artificial intelligence (GenAI) is no longer a futuristic promise. It is fundamentally reshaping value chains, knowledge production logics, and social interactions. The “Generative AI Outlook Report,” published by the Joint Research Centre (JRC) of the European Commission in 2025, emphasizes this. “This is not merely a software advancement, but a paradigmatic shift in how humans produce, consume, and judge information.”
Generative AI models—particularly LLMs (Large Language Models)—offer unprecedented capabilities for writing, coding, simulating, visualizing, or dialoguing with digital systems in natural language. While the promises are enticing, the geopolitical, economic, and social challenges they raise require clear-eyed analysis. The European Union is attempting to establish itself as a balancing force, between innovation and regulation. Entrepreneurs have a crucial role to play in this.
An Amplification Capability at the Heart of Entrepreneurial Dynamics
Generative AI acts as an efficiency multiplier, both in creativity and automation. For startups, SMEs, or innovative companies, it becomes a strategic lever. The JRC report notes: “Entrepreneurs already perceive generative AI as a co-pilot for design, prototyping, customer relations, and experimentation.”
Concrete examples illustrate this dynamic:
- The startup Beink uses image and text generation models to produce customized marketing design proposals for luxury brands in just a few hours.
- At Linkup.io, generative AI powers recruitment assistants capable of cross-referencing CVs, LinkedIn profiles, and job postings in natural language.
- Saagie, a French data ops company, develops connectors with LLMs to automatically generate technical documentation and incident summaries.
This movement is not limited to digital sectors. In healthcare, industry, or logistics, use cases are multiplying: generation of medical reports, synthesis of technical standards, development of digital twins.
Structural Risks for European Societies
But this amplification capability comes with profound risks. The report warns against uncontrolled dissemination: informational hallucinations, model opacity, value capture by dominant players.
“Generative AI is currently largely dominated by a handful of global players, often outside Europe, using closed, non-auditable models trained on opaque data.” (JRC, 2025)
The resulting technological dependence threatens the strategic autonomy of states and businesses. Added to this are complex ethical issues: reproduction of biases, risks of information manipulation, massive disinformation, digital identity falsification.
In this context, integrating a vigilance approach from the design and deployment of solutions becomes an entrepreneurial imperative.
European Ambition: Towards Enhanced Regulation
The European Union, with the AI Act, the Data Strategy, the Digital Services Act, and the Copyright Directive, is attempting to define a balanced regulatory framework. The objective: to guarantee security, robustness, transparency, and fairness of AI systems.
But the JRC report reminds us that regulation alone is not enough. It must be conceived as a catalyst for ethical innovation, not as a brake.
“European regulation cannot be limited to putting out technological fires. It must build conditions of possibility for innovation consistent with democratic values.”
Thus, Europe actively promotes alternative models to American or Chinese giants. Among them:
- Mistral AI (France), which develops high-performance and compact open-source LLMs.
- Aleph Alpha (Germany), committed to algorithmic transparency and user control.
- Lumière (Belgium), a consortium aiming for sovereign training of multilingual models.
These initiatives still need to pass the test of large-scale industrialization. Entrepreneurs can play an accelerating role by integrating these sovereign building blocks into their architectures.
SME Integration: Between Promises and Constraints
Despite their potential, generative AI tools sometimes remain out of reach for small organizations. License costs, API complexity, talent shortage… The obstacles are real.
The report emphasizes the importance of active public policies:
- Resource pooling through AI Factories,
- Simplified access to pre-trained models via European marketplaces,
- Specific training for micro-enterprises and SMEs.
Some initiatives are moving in this direction:
- The Horizon Europe program funds AI experiments in industry.
- The Leonardo supercomputer, based in Italy, supports the training of European models.
But the adoption gap remains significant. Only 4% of European SMEs reported using a form of generative AI in 2024 (source: Eurostat). The challenge is therefore also cultural: transforming the perception of AI from gadget to strategic tool.
Ethics Embedded in Innovation
European singularity lies in its “techno-humanist” approach. The JRC defends a vision where AI is aligned with fundamental rights, human dignity, explainability, and proportionality.
“Europe must embody another path: neither Californian digital ultraliberalism nor Chinese techno-authoritarianism, but governance by principles and dialogue.”
This ethical ambition is not incompatible with entrepreneurial agility. On the contrary: it constitutes a competitive advantage in a world where consumers demand more transparency, accountability, and sustainability.
Some players have understood this:
- Openvalue, a French consulting firm, integrates ethical audits from the AI agent design phase.
The Skills Challenge: Training Tomorrow’s Leaders
A consensus is emerging: generative AI disrupts the skills necessary for organizational leadership. Far from simple technical acculturation, it’s a new grammar of work that must be integrated.
The report emphasizes the urgency of:
- Equipping managers to understand the limitations and potentials of LLMs,
- Integrating AI into initial training at business schools, engineering schools, and design schools,
- Educating hybrid profiles capable of bridging technical, regulatory, and business issues.
Pioneering initiatives are developing:
- The AIxDesign program in Amsterdam trains designers in responsible use of generative AI.
- HEC Paris now integrates a common AI core into its Grande École program.
- Université Côte d’Azur offers interdisciplinary modules on AI, ethics, and organizational transformation.
Entrepreneurs have an exemplary role to play here: creating learning environments, promoting team skill development, and recruiting based on cognitive diversity criteria.
Participatory Governance for Trustworthy AI
The governance of generative AI cannot be exclusively technocratic. The JRC calls for a multi-stakeholder dynamic involving researchers, citizens, businesses, and public decision-makers.
“The development of use cases must be co-constructed with society, based on transparency, contestability, and inclusivity.”
This requires:
- Fostering open, territorialized experiments with shared feedback,
- Supporting citizen deliberation initiatives (workshops, conventions, ethical hackathons),
- Involving professional associations, unions, and local authorities.
Some territories are already experimenting with this. In Barcelona, a citizen committee on AI guides technological investment priorities. In Tallinn, an algorithmic oversight framework was co-written between the state and local businesses.
European Entrepreneurs: Scouts of Responsible Transition
The report’s message is clear: generative AI is a strategic space that Europe cannot delegate. Yet entrepreneurs are at the frontier of this transition: they test, adapt, transform, and model.
Through their technological choices, management practices, and ethical trade-offs, they embody a certain vision of European innovation.
It is now their responsibility to: favor technological choices aligned with democratic values, and share their feedback to foster collective intelligence. And also actively contribute to the emergence of a European ecosystem of trust.
Shared, Embodied, Proactive Sovereignty
Generative AI represents both a lever for geopolitical influence, a catalyst for economic transformation, and a civilizational challenge. The European Union, by embracing an ambitious regulatory line, proposes a third way.
But this path will only be viable if it relies on clear-sighted, agile, ethical entrepreneurs connected to real-world uses and common values.
“Digital sovereignty is not just a matter of infrastructure: it is a cultural, strategic, profoundly human project.”
It is in this perspective that the future of European generative AI will take shape—not in closed laboratories, but in the open, informed, and demanding practices of its entrepreneurs.
Bibliography
- Abendroth Dias, K., Arias Cabarcos, P., Bacco, F.M., Bertoletti, A., Navajas Cawood, E. et al., Generative AI Outlook Report — Exploring the Intersection of Technology, Society and Policy, Joint Research Centre, Publications Office of the European Union, 2025. https://data.europa.eu/doi/10.2760/1109679
- European Commission, AI Act Proposal, Official Journal of the European Union, 2021. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52021PC0206
- European Commission, European Data Strategy, European Commission, 2020. https://ec.europa.eu/info/strategy/priorities-2019-2024/europe-fit-digital-age/european-data-strategy_en
