OVHcloud aims to become Europe's second-largest LLM provider, trailing only Mistral AI, by training its own frontier AI models. CEO Octave Klaba announced at VivaTech that the cost for training such models has decreased significantly, now estimated at 150-200 million euros. The company has already pre-trained a model using Europe's fastest supercomputer, Jupiter, with plans to open-source it once viable and to avoid using client data for training. Meanwhile, Meta is experiencing a leadership change within its "AI for Work" initiative as Emily Dalton Smith departs after a brief tenure, marking another senior exit from the company's AI division. Separately, the US government has ordered Anthropic to cease operations of its Claude Fable 5 and Claude Mythos 5 models, citing a breach in Fable's cyber guardrails, which Anthropic disputes as minor flaws already present in competing models. AI
IMPACT This cluster highlights diverse AI industry movements, from European companies challenging established players to leadership shifts and regulatory scrutiny impacting model deployment.
RANK_REASON Cluster covers a major European tech company's AI ambitions, a significant leadership change at a major AI player, and regulatory action against a frontier AI model provider.
Read on Mastodon — mastodon.social →
- AI for Work
- Anthropic
- Claude Fable-5
- Claude Mythos 5
- Emily Dalton Smith
- Meta
- Mistral AI
- Octave Klaba
- OVHcloud
- US government
- VivaTech
AI-generated summary · Google Gemini · from 3 sources. How we write summaries →