Synthesia’s $4B moment

Here’s what’s worth your time today.

[2 min read]

UK-based AI video startup Synthesia has raised a $200M Series E at a $4B valuation, nearly doubling its worth in a year as demand for AI-powered corporate training continues to climb. The company, which crossed $100M in ARR in 2025 with customers like Bosch, SAP, and Merck, is also running a structured employee secondary sale at the same valuation, letting early staff cash out without going public. Backed again by GV, Accel, Kleiner Perkins, NEA, and NVentures, Synthesia now plans to push beyond avatars into AI agents that let employees interact with company knowledge through role-play and Q&A, signaling how enterprise AI is shifting from content creation to active workplace assistants.

Supported by Vanta

With customer expectations rising and compliance needs shifting, keeping up can be daunting. Vanta helps fast-moving startups and security teams get audit-ready fast and stay continuously compliant, turning compliance into a deal accelerator, not a blocker.

Join the demo to learn how Vanta can help you:

Watch now.

[1 min read]

Nvidia is investing $2 billion in data center provider CoreWeave, deepening a partnership that sits at the heart of the AI infrastructure boom. The deal values CoreWeave shares at $87.20 and sent the stock up more than 10% pre-market, signaling continued confidence in demand for AI compute. Nvidia says CoreWeave’s “AI factory” expertise is critical to meeting surging demand for its chips, underscoring that despite growing questions about circular AI deals, spending across the stack is still accelerating.

[7 min read]

Anthropic CEO Dario Amodei says the world is sleepwalking into a dangerous phase of AI development, warning that super-human systems could cause civilization-level harm without faster, more serious intervention. In a new essay shared ahead of publication, Amodei argues that within a few years AI could outperform humans across most fields, creating what he calls a “country of geniuses in a datacenter.” He flags risks ranging from mass white-collar job disruption to biosecurity threats and AI-powered authoritarianism, and says governments and AI companies are not treating these scenarios with the urgency they demand. His core message is blunt: the technology is advancing faster than society’s ability to control it, and honesty about the risks is now essential.

[2 min read]

Anthropic is rolling out interactive apps inside Claude, letting users connect workplace tools like Slack, Figma, Canva, Box, and Clay directly within the chatbot. The feature allows Claude to access logged-in services so users can send messages, edit designs, generate charts, or pull files without leaving the interface. Available to paid plans only, the move brings Claude closer to being a true work hub and sets it up to pair with Anthropic’s new Cowork agent, which can handle multi-step tasks across connected tools, though users are warned to be careful with permissions as agentic systems grow more powerful

[2 min read]

Microsoft has rolled out the second generation of its in-house AI chip, Maia 200, and paired it with new software designed to chip away at Nvidia’s biggest advantage with developers. The chip is now live in a Microsoft data center in Iowa, with another site planned in Arizona, as the company accelerates efforts to rely less on Nvidia hardware. Alongside the silicon, Microsoft is backing Triton, an open-source programming tool positioned as an alternative to Nvidia’s CUDA, long seen as the glue that locks developers into Nvidia’s ecosystem. The bigger signal isn’t raw performance. It’s that cloud giants are no longer just competing on chips, but on the software layers that decide where AI workloads actually run.