8 articles
Mistral AI, France's leading AI lab, is a customer of Nvidia's US-built infrastructure — about the extent of hard European detail in this global agentic-AI story. A Google Cloud-linked adoption survey, plus Manulife, Box and C3.ai filings, sketch the trust picture, but Via News found the survey source failed nearly every reliability check it ran.
Google has loosened its AI watermark rules and OpenAI has lost another senior executive, but the more concrete story for European business is elsewhere: Spain's Multiverse Computing and Germany's Black Forest Labs are posting benchmarked efficiency gains while some of the loudest US capex and opportunity claims trace to low-reliability sources.
A Grenoble-based chipmaker's new AI-networking tie-up with Bull lands the same week SOXX has fallen 25% from its highs — evidence that the divergence between deal-making and equity sentiment now runs through Europe too, verified against original filings and press releases.
A late-July rout wiped tens of billions off AI-chip stocks even as SK Hynix, Nvidia and Washington doubled down on the buildout. Via News traces the numbers — from a Grenoble networking deal to Vicor's audited revenue — to separate the correction from the cycle.
Nearly all enterprises plan to run AI agents within two years, but the survey behind that figure checked poorly against primary sourcing. What holds up: European regulated-industry logic — auditable, reproducible, hallucination-resistant automation — is where vertical AI is actually monetizing first.
Brussels is reviewing its Code of Practice on Transparency of AI-generated content just as a Spanish deep-tech firm demonstrates the case for it, cutting AI inference latency 48.6% without sacrificing accuracy, while press releases from three US AI ventures cited here are shown by Via News's own reliability tracking to hold up only one time in five.
