Investments 💶

🔌 “Plug baby Plug” Macron unveils a total of €109 billion in planned private investment in AI infrastructure

This eye catching figure includes €30-50 billion from the UAE, €20 billion from Brookfield Asset Managers, €6 billion from Amazon, and additional investments from firms like Apollo Asset Management. Macron likened the sum of these investments to the $500 billion Stargate AI investment program in the United States. These new funds in France will primarily be used for infrastructure projects—like the construction of 35 new data centers.

The investment will also make France the single largest market in Europe for Brookfield and their datacenter partner Data4. To support this position, Brookfield is also opening a new office hub in Paris. (Forbes)

💶 €200 billion announced for EU AI investments

The mix of public (€50 billion from the EU’s InvestAI initiative) and private funds aims to deploy this large pool of capital over the next five years. €20 billion from this amount has been allocated to fund the construction of four new massive computing hubs dubbed ‘GigaaAI factories’ The EU—adding to the seven planned sites announced in December. (Le Monde)

🇫🇷 Bpifrance will invest €10 billion into French AI by 2029

The investment will be split into “direct” and “indirect” categories, with direct investment going to AI startups and the indirect investment going to support AI infrastructure. "Thanks to our investments and continued support, we are strengthening France's position as a global player in this strategic field," said Bpifrance CEO Nicolas Dufourcq. (CB News)

👨‍🔬 Xavier Niel’s Iliad Group to allocate €3 billion to AI infrastructure, research, and applications

As part of the announced allocation, Iliad Group subsidiary OpCore will invest €2.5 billion in hyperscale data centers backed by a partnership with private equity firm InfraVia. Another part of the funding is being used to purchase 5,000 GPUs to establish “the largest AI compute capacity available in the commercial market in Europe.” (Iliad Group)

Infrastructure 👷‍♀️

🏗️ Swedish cloud startup Evroc will build their second hyperscale data center in Mougins, 500 miles south of Paris

The planned 96-megawatt facility will fit 50,000 GPUs and require an estimated €4 billion to complete. Evroc intends to have the datacenter built and running in 2025. (Sifted)

🆕 Mistral will open an AI cluster in Essonne

The scalable data center will be equipped with the latest chips, powered by decarbonized energy, and is expected to be operational in the coming months (LinkedIn)

Partnerships 🤝

⚔️ Mistral partners with defense tech startup Helsing

Together, the companies aim to develop next-generation models for military operations. Helsing, which is active in France, Germany, and the UK, develops AI-enabled attack drones—of which 4,000 units have already been sold to Ukraine. (Le Monde)

🟧 Orange and Mistral AI join forces with new infrastructure partnership

Orange and Mistral AI have formed a strategic partnership to enhance their AI capabilities in Europe. The collaboration includes integrating Mistral AI’s technology into Orange’s network infrastructure and business offerings and expanding AI-driven services. (Orange)

📱 Mistral le Chat Pro will be available for Free’s 15.5 million subscribers

With this new partnership, all Free subscribers will get 12 months of free access to the premium version of Mistral’s AI assistant. (Iliad Group)

🔎 Bouygues Telecom will offer Perplexity Pro for free

Bouygues Telecom is partnering with U.S. startup Perplexity to offer their AI-powered search engine, Perplexity Pro, free to all subscribers for one year (normally $240). The move marks Bouygues' first major AI initiative. (Usine Digitale)

💼 France Travail and Mistral AI partnering on job search services

Mistral AI has partnered with France Travail to enhance job search services using generative AI. The agency will deploy ChatFT, a chatbot for writing assistance, and MatchFT, an AI tool that contacts pre-selected candidates via SMS. Tested in 87 agencies, the system aims to streamline recruitment. (France TV)

Organizations, Governments, and NGOs 🏛️

📝 60 countries sign the Paris AI Summit Declaration

Over 60 countries signed an agreement promoting ethical, transparent, and inclusive AI development while addressing digital divides and market concentration while notable holdouts the US and UK declined to join. (Computer Weekly)

♻️ The Coalition for Sustainable AI is launched

This new global initiative is aimed at aligning artificial intelligence development with environmental sustainability. Initiated by the French government in collaboration with the United Nations Environment Programme (UNEP) and the International Telecommunication Union (ITU), AI leaders like Photoroom, Hugging Face, and Cohere are part of the initial signatories. (Sustainable AI Coalition)

⚡ France to allocate nuclear energy for data centers

As part of the push to build critical AI infrastructure, France has promised to allocate power from its 57 nuclear reactors, with a goal of 250 megawatts powering data centers by 2026. I’ve previously written about how France’s high availability of nuclear energy can be an advantage in the global AI race. Josh Parker of NVIDIA’s seemed to back this up at the Summit saying “Power is the number one consideration for getting access to AI and bringing our systems online.”

Some have raised concerns over questionable capabilities to quickly build out the connective systems that link that energy to data centers. However, EDF has already identified four promising sites with existing electrical connections and availability for up to 2 gigawatts of power. (WSJ)

🛟 Introducing Current AI, the first foundation dedicated to the AI commons

Current AI, a nonprofit backed by €400 million in initial funding from the French government, AI Collaborative, Google, Salesforce, and others will aim to enhance AI data quality and accessibility. The initiative aims to raise $2.5 billion over five years, focusing on open data, transparency, and public-interest AI. The nonprofit’s first project, ROOST, is centered around developing AI safety infrastructure. (Current AI)

📈 AI Champions Initiative plans to advance AI in Europe

More than 70 organizations across Europe have joined the AI Champions Initiative, which aims to establish Europe as a global leader in AI by “mobilizing talent and capital, accelerating AI adoption in key industries, and fostering a competitive technology ecosystem to enhance economic sovereignty and resilience.” Airbus, Doctolib, Renault, and many other leading French names have signed on in support of the initiative. (AI Champions)

🇫🇷 France to equip 5.7 million public employees with a sovereign AI

The French government plans to roll out access to a sovereign generative AI tool in the coming months to streamline administrative tasks such as research, drafting letters, and writing reports. Employees will also receive AI training in the coming weeks to facilitate adoption. The state will issue a call for proposals to identify AI solutions with Mistral mentioned as a potential pick. (Banque Des Territoires)

Safety 👮‍♂️

🦺 Inauguration of INESIA: the French "AI Safety Institute"

France is launching a new national institute dedicated to evaluating AI security, bringing together key research and cybersecurity institutions to assess AI’s risks and ensure safe development. INESIA joins a network of 10 AI Safety Institutes worldwide—including in the US, UK, Japan, and Canada. (Campus France)

🔔 Anthropic takes a different tone on the week, focused on safety

In a statement from CEO Dario Amodei, Anthropic urged countries and companies to focus on three key areas to create the best future outcomes in a world heavily impacted by AI. (Anthropic)

1. Ensure democratic societies lead AI development.

2. Focus international conversations on AI to address security risks from the technology.

3. Enact policy to ensure everyone shares in the economic benefits of AI, not just a group of geniuses in a datacenter.

Education 🎓

👩‍🎓 AI classes will be mandatory for schools in France

Beginning in September, middle and high schools in France will offer online AI training for students. The courses will be mandatory for 8th and 10th graders and optional for other students. (L’Etudiant)

🍎 AI hubs to train 100,000 people on AI each year

According to data from the government, France currently trains 40,000 students and professionals on AI each year. With the creation of 9 “AI Cluster” training hubs and additional funding, the government plans to increase this number to 100,000 each year. (Elysee)

Drama 😱

🇺🇸 US VP touts US AI leadership and bashed EU regulations

In his talk at the summit, U.S. Vice President JD Vance warned against excessive AI regulation and criticizing Europe's Digital Services Act for burdening American tech companies. He also contrasted the U.S. approach with China's AI policies and heralded the US as the current and future leader in AI development. (Le Monde)

✂️ Revisions made to the European AI Act

The European Commission has withdrawn the AI liability directive following criticism from US Vice President JD Vance at the AI Action Summit in Paris. The directive, which aimed to adapt civil liability rules to AI and ensure victims could claim compensation for harm caused by AI, had already lost momentum prior to Vance’s speech. (Euractiv)