ENFR
8news

Tech • IA • Crypto

TodayBriefingVideosTop 24hCryptoArchivesFavoritesTopics

VivaTech and Osez l'IA reveal the real crisis of French AI

5/10
AIRenaud DékodeJune 23, 2026 at 02:15 PM20:11
Audio player
0:00 / 0:00

TL;DR

The latest edition of VivaTech and France’s reinforced OseIA plan are criticized for prioritizing image and communication over concrete technological progress and effective AI adoption.

KEY POINTS

Shift from innovation to branding at VivaTech

The event appears to have evolved into a showcase dominated by large corporate “villages,” where major brands, regions, or countries host clusters of startups. These startups are relegated to minimal, standardized booths with limited visibility, often reduced to a small screen and a nameplate. The focus has shifted toward highlighting sponsor logos rather than showcasing the substance of innovation.

Decline in tangible technological demonstrations

Compared to previous editions, there is a noticeable absence of cutting-edge prototypes such as autonomous vehicles, drones, or experimental mobility solutions. Only isolated examples were observed, suggesting a reduced emphasis on breakthrough technologies and hands-on demonstrations.

Limited presence of research and public sector actors

Public research institutions and exploratory science, once central to the event’s appeal, are now largely absent. Aside from a modest CNRS presence, there is little representation of scientific laboratories or forward-looking research initiatives. Public services such as law enforcement or administrative innovation, previously visible, are also missing.

Controlled communication and lack of open exchange

Corporate participation is increasingly restricted to top-down presentations and staged conferences. Opportunities for direct questioning, demonstrations, or independent interviews are limited, reinforcing the perception of a tightly managed communication environment rather than an open innovation forum.

Event perceived as a communication platform

The overall transformation suggests a move away from collaborative exploration toward a platform focused on visibility and reputation. The emphasis on large-scale branding and curated messaging has reduced opportunities for meaningful debate, comparison of technologies, and collective intelligence.

Concerns over the OseIA national strategy

The French government’s reinforced OseIA plan aims to accelerate AI adoption, targeting 100% of large companies, 80% of SMEs, and 50% of small businesses by 2030. However, the measures announced raise questions about their practical effectiveness in reaching these ambitious goals.

Reliance on consultants and commercial actors

The expansion of “AI ambassadors” includes primarily service providers and consulting firms, rather than independent educators or practitioners. This approach is seen as favoring commercial interests, with public funding potentially supporting private sales activities rather than broad knowledge dissemination.

High-cost diagnostics with limited scale

A key measure involves AI diagnostics costing €10,000 for approximately eight days of consulting, with 40% publicly funded. A total budget of €9.6 million would support around 2,400 companies, a small fraction of the millions of businesses in France, raising concerns about scalability and impact.

Additional programs with limited reach

Complementary initiatives, such as “AI accelerators,” are expected to support fewer than 100 companies with deeper guidance. While more comprehensive, these programs remain insufficient relative to national adoption targets.

Critique of strategy effectiveness

Critics argue that the current approach emphasizes symbolic action and communication over pragmatic solutions. Alternatives suggested include large-scale, accessible training programs and direct knowledge sharing to accelerate adoption more efficiently and inclusively.

CONCLUSION

Both VivaTech’s evolution and the OseIA plan highlight a broader concern that AI policy and innovation ecosystems risk being driven by visibility and structure rather than tangible, scalable impact.

Full transcript

More from AI