Anthropic’s slowing growth could be the AI trade’s latest red flag
THE SO WHAT
Public-market AI sentiment is starting to care about revenue durability, not just model demos — Anthropic’s cooling growth ahead of a listing will feed that repricing. If you’re selling into the AI stack, assume capital is getting more selective and tighten your path from infra spend to contracted revenue.
READ THE SOURCE
MORE FROM THE WIRE
8 states are suing OpenAI, saying ChatGPT harms kids. Here are their demands, from data deletion to halting AI testing.
State AGs are moving from hearings to hard remedies—data deletion, age checks, and limits on advanced testing are now on the table. Anyone shipping consumer-facing AI should assume child-safety rules will harden quickly and design data retention, eval, and rollout processes as if audits are coming.
Applied AIJPMorgan's Dimon Sees Benefits From AI, Warns of Risks
When the largest U.S. bank CEO talks up AI upside and risks in the same breath, boards hear both mandate and constraint. Financial services teams should expect faster budget approval for AI pilots—paired with tighter model risk, compliance, and operational-resilience gates.
Applied AIRicursive Intelligence is betting on AI-designed chips to help solve the compute crisis
If AI can meaningfully automate chip design loops, the bottleneck in the compute crisis shifts from human layout cycles to capital and fabs. Foundries, EDA vendors, and large AI buyers should be exploring co-design pilots now — whoever closes that loop first gets leverage on cost and time-to-silicon.
Applied AICapsem – a Rust hypervisor sandboxing AI agents in air-gapped Linux VMs
Sandboxing AI agents inside Rust-based, air-gapped VMs treats agent execution like untrusted code by default. If you’re experimenting with agentic systems that touch sensitive systems or data, this is the security posture to benchmark against—not running agents directly in your production environment.