The AI Action Summit kicks off today, bringing together world leaders eager to claim the title of AI pioneers. But behind the stage, France’s AI strategy is at a critical juncture.

AI is not just another industrial revolution—it’s moving at a speed that compresses decades into mere months. If the Industrial Revolution gave the UK dominance for decades, AI is unfolding at light speed—and France risks being left behind. Companies and individuals too.

1️⃣ AI Sovereignty or Just Renting Compute Power?

Yesterday, President Macron has announced €109 billion in AI-related investments—but here’s the problem:

• €30-50 billion will go to data centers.

• Who benefits? The UAE, a Canadian fund, and likely NVIDIA.

• Where are France’s or EU’s AI chips? The EU is pushing semiconductor independence, but France remains critically dependent on U.S. and Chinese chipmakers.

🚨 If you don’t own the hardware, you don’t own the AI revolution.

Despite having clean, nuclear-powered energy, French businesses still pay the same electricity prices as other EU nations that abandoned nuclear. If AI infrastructure is built on imported chips, foreign-funded data centers, high energy costs, and that all profits goes to UAE or Canada, can we really call this AI sovereignty?

France must invest in its full AI value chain—compute power, foundational models, cloud infrastructure, unique business offers, and a ready workforce —or sovereignty will remain an illusion.

At their level, companies should also do same and go beyond using “ChatGPT” or “Le Chat”.

2️⃣ France Needs Fewer Startups, More Global Leaders

President Macron recently claimed France is the 5th AI power globally—but on what basis?

• The number of AI startups has exploded: 2018: 270 startups, 2025 (early last week): 700 startups. End of last week: 1,000 startups

• Are these real AI innovators or just companies rebranding themselves as “AI startups” to attract investment?

• How many French AI companies are scaling globally?

🚨 Without scale, France will remain an AI talent farm—training engineers who get poached by the U.S.

Instead of counting startups, France must build global AI champions. This means:

✅ Mergers & alliances instead of fragmentation.
✅ More late-stage funding to help AI startups scale.
✅ Focusing on deep-tech innovation, not just SaaS applications merely using a LLM

If France doesn’t consolidate its AI sector, it will always be playing catch-up.

3️⃣ From Research to Reality: Where Are the Industrial Applications?

France is home to some of the best AI researchers in the world, but has failed to translate research into large-scale industrial applications.

• China had an AI plan in 2015. Today, it leads in autonomous systems, AI-driven manufacturing, and surveillance tech.

• France had an AI plan in 2018. Where are the results?

🚨 A full audit by the Cour des Comptes is needed—where have France’s AI investments gone since 2018?

While president Macron is now pushing “frugal AI” (energy-efficient models), how many French companies are truly leading in distilled models, edge AI, or model compression? And it this enough?

✅ A strategy without execution is just a wish list. France must bridge the gap between AI research and real-world applications. How? Part of the solution is probably interoperability and standards. France’s biggest power is probably still its consumer market. By setting the proper standards, France and EU can drive a desired future.

4️⃣ AI & Cybersecurity: Still an Afterthought?

Cyber threats powered by AI are already here:

• Deepfake fraud and voice impersonation scams are skyrocketing.

• Data leaks make it a heaven for criminals

• AI-powered hacking tools are automating cyberattacks at scale.

Yet France’s AI and cybersecurity sectors still operate in silos.

🚨 If AI isn’t designed with cybersecurity from the start, France will be vulnerable before it even scales.

The EU AI Act is setting global regulatory benchmarks, but that’s just compliance.

✅ France must lead in AI security innovation—not just follow regulations.

✅ AI and cybersecurity must be integrated from the business specification phase, not added as a patch later.

✅ Companies must treat security as a core feature, not an afterthought.

🔎 If AI isn’t secure, and fully trusted, it is a liability.

5️⃣ AI Will Increase Inequality—Unless France Takes Control

President Macron insists AI “will not replace humans, but will give them more quality time.” That’s optimistic, but the reality is harsher:

• AI is already automating white-collar jobs.

• Workers who don’t learn to use AI effectively will fall behind.

• Education must pivot fast—will France train AI creators, or just passive AI consumers?

🔎 The biggest risk isn’t job loss—it’s a widening skills gap.

✅ France must invest in AI education—from K-12 to workforce reskilling.
✅ AI shouldn’t just be something people use—it should be something they create.

The next generation must be trained to build, fine-tune, and control AI systems—not just be at the mercy of them.

🚀 Final Call: France Must Act, Not Just Talk

🇫🇷 France’s AI future depends on bold choices. Instead of self-congratulation, leaders at the AI Action Summit must:

✅ Invest in scaling AI startups into global companies.

✅ Turn AI research into real industrial applications.

✅ Build AI with cybersecurity from the start.

✅ Secure AI hardware & cloud infrastructure—don’t just rent it.

✅ Equip the next generation to be AI creators, not just consumers.

💡 The AI race isn’t about who gives the best speeches—it’s about who executes best. France still has a shot—but time is running out.

For the rest of us, leaders and experts at companies, organizations of all sizes, it’s more than time to take control of AI. Or it will control you.

Ressources

Well, the president also thinks that AI will not replace humans, but will allow humans to have more quality time. Let’s hope he’s right. I’m not an AI, but as promised this week on LinkedIn, here are the ressources that will help you with tacking another pressing issue, DORA security compliance. And sleep better.