Quoi de neuf en intelligence artificielle – juin 2025

June 2025 brought a flurry of breakthroughs, regulatory shifts, and corporate strategies strengthening the global AI narrative—from national policy moves to the rise of agentic AI assistants and major AI infrastructure investments.

📌 Major Developments & Policy

U.S. legislative friction threatens AI competitiveness

Le nouvellement décédé «Une grosse et belle facture» includes cuts to energy incentives and social programmes, potentially raising utility costs and hampering America’s capacity to support energy‑intensive AI data centers. Observers warn it may erode U.S. AI leadership amidst global competition. 

Lawmakers push to ban Chinese AI tools from U.S. agencies

A bipartisan bill was introduit to restrict federal agencies from using AI systems developed by strategic competitors like China, citing risks to national security and technological sovereignty. 

EU’s wave of AI gigafactory proposals

L'Europe a reçu 76 offres à travers 60 emplacements from tech and industrial firms to establish AI gigafactories—large‑scale facilities laid out to house ~100,000 next‑gen AI chips. This aligns with a €20 billion EU investment in digital infrastructure. 

🏢 Corporate Strategies & AI Agents

Métaformes « Superintelligence Labs »

Meta a annoncé le lancement de son Laboratoires de méta-superintelligence (MSL), underpinned by leadership from veteran AI figures like Alexandr Wang and Nat Friedman. This marks an aggressive push into open‑source “Llama” models and next‑gen AI research. 

Apple considers switching Siri to Anthropic or OpenAI

Apple is exploring integrating OpenAI or Anthropic models to power Siri, signaling a major shift from in‑house AI. While no final decision is made, it reflects Apple’s broader strategic AI reshuffling. 

🤖 AI Agents & Consumer Innovation

L'Inde lance l'assistant agent « Kruti »

Ola Krutrim dévoilée Kruti, a multilingual AI agent capable of handling everyday tasks—such as booking cabs or ordering food—in 13 Indian languages. Designed to understand context and remember user preferences, Kruti demonstrates the emergence of agentic AIs in emerging markets.

Humanoid AI robots play real-world soccer

Le RoBoLigue in Beijing featured its first 3-on-3 soccer match entirely played by humanoid robots. Driven by reinforcement learning and autonomous perception, this event underscores the scaling of physical AI applications. 

Amazon atteint le million de robots déployés

Les centres de distribution mondiaux d'Amazon fonctionnent désormais sur 1 million de robots, propulsé par le nouveau Flotte profonde generative AI model—improving fleet efficiency by 10% and accelerating package delivery. 

🏥 Government & Public Sector AI

FDA launches “Elsa”, its first AI tool

On June 2, the U.S. Food & Drug Administration introduit Elsa, a generative AI assistant designed to support regulatory work, data review, and decision-making à travers departments. 

Enquête : 1 étudiant sur 5 utilise ChatGPT pour obtenir des conseils de carrière

Une récente enquête sur les débuts de carrière montre 20% des étudiants are consulting tools like ChatGPT for job search and career planning support—signalling AI’s growing influence on youth employment decisions. 

⚙️ Research, Sustainability & AI Infrastructure

IBM-led workshop outlines AI sustainability roadmap

Un atelier interdisciplinaire intitulé “Greening AI with Software Engineering” proposed research agendas to reduce the carbon footprint of AI systems—focusing on energy benchmarking, architectural standards, and runtime adaptivity. 

Le supercalculateur japonais ABCI 3.0 sera opérationnel début 2025

Le Japon ABCI 3.0 AI infrastructure (launched officially January 2025) is now fully operational, delivering over 6 exaflops of performance à travers thousands of Nvidia H200 GPUs. It’s poised to drive next-gen AI research and training. 

📊 Market & Industry Signals

• AI market growth slows but remains robust: la croissance projetée atteindra 26 % en 2025, contre 33 % en 2024, mais cela indique toujours une expansion forte et soutenue.

• Morgan Stanley and McKinsey bullish on quantum‑AI synergy: expect quantum computing and AI to combine powerfully in sectors like finance, biotech, and materials science.

✅ Résumé : ce que nous dit le mois de juin

June 2025 reflects a maturing AI landscape with:

• Sharpening geopolitical fault lines in AI policy and infrastructure.

• Rising prominence of agentic AI both in consumer markets and enterprise operations.

• Increasing government deployments of AI tools in public service.

• Greater emphasis on sustainability and ethical practices in high-performance AI.