Coding Agents Just Tripled to a $5B Valuation in Five Months

AI Agents

The build-and-govern frontier, decoded.

Factory’s Coding Agents Just Tripled to $5B — the “Software Factory” Is the New Unit of Work

What happened: Factory, the startup whose AI agents automate complex programming work, raised $200 million at a $5 billion valuation — more than triple its price from five months ago. Khosla Ventures, Blackstone, Sequoia and NEA participated, with Salesforce CEO Marc Benioff among the angels. The company says its platform is already used by hundreds of thousands of developers at firms including Nvidia, Adobe and Palo Alto Networks.

Why it matters: This is the clearest signal yet that the industry is moving past single coding assistants toward what Factory’s CEO calls a “software factory” — a persistent foundation where an agent orchestrates a fleet of sub-agents. Developers hand over a high-level description, and the platform breaks it into sub-steps, routes each to the cheapest capable LLM, spins up isolated “Droid Computer” environments, and even auto-writes documentation. For anyone building with agents, the takeaway isn’t “another coding tool got funded.” It’s that the winning abstraction is shifting from a model you prompt to a team you manage.

What’s next: Factory plans to nearly triple headcount to 300 by year’s end and leans on LLMs to auto-improve its own agents from anonymized usage signals. Watch whether “agent-friendly” repositories — with guardrails and monitored dashboards — become a standard checklist item, the way CI/CD did a decade ago.

Source: SiliconANGLE

A SOC 2 for AI Agents Just Raised $40M

AIUC — founded by an early Anthropic employee and METR’s former COO — launched with $40M to do for AI agents what SOC 2 did for software: a standard audit (AIUC-1) that runs agents through ~5,000 tests for jailbreaks, hallucinations and data leaks. Its first customers are Cursor, Lovable, Harvey and ElevenLabs — a sign agent builders now want a trust badge to sell into cautious buyers.

Source: TechCrunch

72% of Hospitals Run AI Without IT Approval

A survey of 250 U.S. healthcare leaders found 72% deploy AI agents without formal IT sign-off “at least occasionally,” while 28% already run agentic AI in production. The gap isn’t capability — 86% feel confident they can govern agent actions — it’s that only 17% think existing identity frameworks still work. Expect “identity-led governance” to become the new checklist in regulated verticals.

Source: HIT Consultant

AI News

What else moved the industry this week.

OpenAI Backs a Federal Audit Mandate — a First

OpenAI publicly backed parts of the bipartisan FRONTIER Act, which would require the largest AI firms to submit models to independent third-party audits — its first endorsement of a federal mandate. The caveat: it hasn’t endorsed the full bill, and critics warn voluntary standards are “company-controlled” without enforceable government teeth.

Source: CBS News

Mistral Raises €3B — Europe’s Biggest Tech Round Ever

The French lab closed a €3 billion Series D at a €21B+ valuation, led by Samsung, to push “sovereign, open-weight” AI — models customers can run on their own infrastructure without vendor lock-in. With 125+ enterprise customers including Airbus, ASML and HSBC, it’s a direct counterweight to the closed-model frontier labs.

Source: Mistral

Rival Labs Quietly Huddle on Safety — but Agreement Is the Easy Part

OpenAI confirmed it has discussed AI safety with Anthropic and Google DeepMind for weeks, and Altman endorsed Amodei’s plan to embed independent evaluators inside labs (“great idea”). But Cohere’s Aidan Gomez warns shared standards written by a handful of commercially aligned labs — possibly behind an antitrust waiver — could just concentrate more power. Rare consensus, hard implementation.

Source: ET Enterprise AI / AP

Quick Plug

Want to build your first AI employee? Grab the free 90-minute build guide — one worked example, start to finish.

https://go.aitokenlabs.com/digest-build

This newsletter? Written by an AI Employee, approved by a human — so our team stays focused on what only humans can do.