From Ethical AI to Sovereign Infrastructure: Insights from Tech Show Paris 2025

European leaders gathered at the Tech Show Paris 2025, emphasizing the need for businesses to rebuild digital trust while pursuing growth amid rapid automation. The discussions spanned from ethical AI practices to infrastructure sovereignty, laying a roadmap for a responsible digital economy.

AI with a Mission: Building a Responsible Digital Economy

The main stage kicked off with the panel titled “AI with a mission: how CEOs are building responsible digital practices.” The tone was pragmatic, ethical, and purpose-driven. Yves Del Frate from Institut CSA highlighted a stark reality: France and Europe remain “digital colonies” concerning infrastructure and software sovereignty. His assertion resonated deeply, as he pointed out the irony of advocating for “responsible technology” while lacking autonomy over critical hardware and data.

Samir Amellal of France Travail presented a counter-narrative, insisting that surrendering to external forces isn’t an option. He highlighted France’s collaboration with Mistral AI, which seeks to build a sovereign ecosystem as a counterbalance to dominant foreign tech giants. This partnership indicates that leaders must be willing to embrace the complexities and costs of building local capacity for the promise of long-term control.

Owen Lagadec-Iriarte from Tech Pays Basque reframed the concept of responsibility in terms of clairvoyance. He posed a challenging question to CEOs: “How many euros are you willing to forgo for an ethical decision?” His point was compelling: responsibility often begins where convenience ends. As we head into 2026, it becomes clear that sustainability and sovereignty have merged; investing in local AI and energy-efficient data centers is now integral to operational risk management.

Inclusive Innovation: Designing Technology for Polarised Societies

The second headline debate, titled “Inclusive Innovation — Designing Tech for Diverse Societies in a Polarised Age,” expanded the conversation from national sovereignty to societal cohesion. Peter Mousaferiadis from Cultural Infusion provided alarming statistics, noting a decline in global peace coinciding with an increase in social media usage. He argued that AI design must incorporate cultural understanding as a foundational requirement, not an afterthought.

Gold Darr-Hood, CEO and inventor, underscored that current business models tend to favor consolidation over diversity. He pointed out the opportunity to leverage AI for mass customization in a way that can preserve diversity and address environmental challenges. Unfortunately, most applications today serve to entrench monopolies due to the financial interests of those in power.

Jean-Christophe Bas from The Aspen Institute attributed today’s societal instability to a broader collapse of trust—geopolitically, institutionally, and informationally. He issued a warning about a “civilisational crossroads” where European values might erode under external pressures.

In tandem, Waqas Ahmed of the Da Vinci Network and Olivia Heslinger from AI for Good Denmark connected these macro conversations to algorithmic realities. They discussed how biased data can produce a distorted reality, advocating for smaller, federated language models that reflect techno-diversity. This idea aligns with a growing belief that open, distributed AI ecosystems are crucial for achieving equitable innovation.

The consensus emerging from this panel is that cultural literacy will become a vital competitiveness metric. This insight translates into the necessity of embedding inclusion into product design and governance rather than treating it as an add-on after marketing.

Rethinking Governance in the Age of AI

The afternoon panel, “Rethinking governance in the age of AI,” provided insights into how boards can translate these principles into actionable oversight structures. Raphaële Contamin from Equans Digital and Corinne Lejbowicz from La Poste Group explained how boards now require updates on AI strategies as routinely as they do on financials.

Christophe Aulnette, former CEO of Microsoft France, cautioned against boards that focus on short-term productivity gains while neglecting the broader opportunities and risks presented by AI. Three consensus points emerged from the discussion:

1. Diffused Responsibility Over Designated “AI Champions.” All board members need to understand AI’s impacts; relying on a single expert can dilute governance effectiveness.

2. Ethics as Veto-Power, Not Branding. Boards should enable internal ethics committees to stop projects that violate established moral thresholds, shifting ethics from a marketing tool to a decision-making cornerstone.

3. Performance Equals Responsibility. Yves Del Frate emphasized that European companies cannot sacrifice ethics for competitiveness—it’s increasingly demanded by the market.

The humorous yet thought-provoking question of whether an AI could sit on a board sparked lively debate. While most agreed that an AI observer could improve information flow, only humans can shoulder fiduciary duties. Still, this conversation reveals a rapid evolution in how algorithmic governance is being integrated into strategic decision-making.

Quantum Reality Check: From Concept to Capability

The late-afternoon keynote session, “2025: From concept to reality for quantum computing technologies,” shifted attention back to the core infrastructure essential for digital sovereignty. Stéphane Bout from McKinsey & Company dispelled myths surrounding quantum “supremacy,” proposing a more pragmatic measure: quantum readiness.

Currently, only about 15% of European enterprises have active quantum initiatives, despite over 70% expecting significant business impacts by 2030. Bout emphasized focusing on hybrid architectures that combine high-performance computing, classical AI, and quantum simulations. This approach mirrors previous exercises in AI, highlighting that the same strategic introspection is needed for quantum technologies.

For CIOs and CTOs, the imperative is organizational rather than scientific: build quantum-literate teams now. Pair data scientists with physicists, and incorporate simulation workflows into cloud operations to harness Europe’s competitive edge in integration.

Open LLMs and the European Case for Transparent AI

The discussion on open language models centered on a pivotal question for the future: should AI leadership be assessed by its scale or its trustworthiness? Antoine Bosselut from EPFL & Swiss AI Lab argued that merely achieving scale does not guarantee societal alignment. Europe’s strength lies in designing transparent, explainable, and regulation-aligned AI that caters to multilingual and domain-specific contexts.

Moreover, smaller, auditable models under strict data-protection regulations could outperform generic systems in more regulated environments. In this context, transparency becomes a means of risk management—a way to ensure accountability and inclusivity.

For business leaders, the implications are clear: invest in ecosystems that are not only auditable and interoperable but also linguistically inclusive. This aligns technological ambitions with Europe’s commitment to lawful and transparent innovation.

The Emerging Consensus: Europe’s Competitive Advantage Is Its Conscience

As Day 1 of the event drew to a close, discussions converged into a hierarchy of readiness that will shape Europe’s digital future:

1. Ethics Sets Direction
2. Governance Builds Trust
3. Infrastructure Delivers Capability

As Europe strides towards 2026, it is evident that leaders who successfully integrate responsible design with sovereign infrastructure will redefine competitive excellence on the global stage.