Daily AI Brief

The company behind the Claude family of models revealed a massive infrastructure commitment: $50 billion will go into building new data-centre facilities in Texas and New York, in collaboration with London-based Fluidstack. This move underscores the escalating compute arms race and the belief that control of hardware pipelines remains key to AI leadership.

Foxglove, which provides data and observability platforms for robotics and "physical AI", announced a $40 million Series B led by Bessemer Venture Partners. The funding will accelerate infrastructure and tooling for robots, logistics, manufacturing and other real-world environments beyond pure software agents.

Atos unveiled its AMOS-AI platform in partnership with Red Hat's OpenShift AI, a hybrid/multi-cloud solution that allows enterprises to train, deploy and manage models securely across on-premises, private-cloud and public-cloud environments — offering regional data-sovereignty, auditability and multi-cloud flexibility.

The analytics provider launched a new product, AI Feedback, designed to turn large volumes of customer-feedback (sales calls, reviews, support tickets) into prioritised insights, user stories, feature-roadmaps and churn-detection, via embedded AI agents. This demonstrates a push into enterprise tooling beyond generative chatbots.

A report by Nymbl Ventures shows that investment in the built-environment sector (construction, infrastructure, digital twins, robotics) reached $4.4 billion in Q3, up 66% year-on-year, with AI-related startups capturing roughly two-thirds of the amount and construction-tech investment tripling. The trend signals broadening of AI investment from software to physical infrastructure.

Theo AI, a legal-tech startup specializing in AI-powered predictive litigation tools, raised $3.4 million in seed funding to expand its platform and help legal teams make data-driven decisions about case outcomes.

Sweet Security, a cloud-security startup focused on AI-model runtime protection, raised $75 million to expand its platform globally and protect enterprises from AI-specific security threats.

Tavus, a startup building multimodal "AI humans" combining text, voice, and facial expressions, raised $40 million Series B to advance its human-computing interface vision and bring more natural AI interactions to enterprise applications.

Atos AMOS-AI is now generally available, enabling enterprises to run generative-AI and agentic workflows in hybrid/multi-cloud environments with auditability and regional sovereignty. The platform addresses critical data governance concerns for enterprises.

Amplitude's AI Feedback tool is live, enabling organizations to turn user-feedback into insights and actionable product plans via embedded AI agents. This moves AI from experimental chatbots to operational product management tools.

An insightful analysis on the built-environment funding surge: how AI and robotics are reshaping infrastructure investment and what it means for industries like construction, infrastructure, real-estate and manufacturing.

A hybrid/multi-cloud platform built for enterprises to train, deploy and manage AI models and agentic workflows securely and regionally. With data-sovereignty and auditability in focus, it is an example of the kind of enterprise-grade tooling that will underpin the next wave of AI scaling.