Weekly AI Intelligence Briefing

Sia Reads
What Matters

Week of March 30, 2026

8 stories tracked

NEWS2026-04-01

OpenAI just raised more than most countries' GDP — Amazon led with $50B, and the $852B valuation puts them one deal away from the trillion-dollar club

9
Read more

OpenAI closed the largest private financing round in history at $122 billion, valuing the company at $852 billion post-money. Amazon led with $50B, NVIDIA and SoftBank each contributed $30B. The company now reports $2B in monthly revenue, nearly 1 billion weekly ChatGPT users, and 50 million paid subscribers.

NEWS2026-03-31

A misconfigured .map file in the npm package exposed 512,000 lines of Claude Code's TypeScript — and the internet kept every line

8
Read more

Anthropic accidentally shipped a 59.8MB source map in Claude Code v2.1.88, exposing nearly 2,000 TypeScript files. The repo hit 84,000+ GitHub stars within days. Anthropic called it human error, not a security breach — but a brief supply chain attack window affected users who installed between 00:21 and 03:29 UTC on March 31.

RELEASE2026-04-02

Gemma 4 ships in four sizes with full Apache 2.0 freedom — no MAU caps, no use restrictions, and agentic workflows running entirely on-device

8
Read more

Google dropped Gemma 4 with models from 2B to 31B parameters, all under a true Apache 2.0 license for the first time. The models support multi-step planning, autonomous action, offline code generation, and audio-visual processing on-device. The 31B dense model competes on raw quality while the 26B MoE variant optimizes for speed.

NEWS2026-03-26

A CMS misconfiguration exposed 3,000 internal assets and confirmed Anthropic's next model tier: Capybara-class, above Opus, with unprecedented cybersecurity capabilities

7
Read more

A configuration error made nearly 3,000 unpublished Anthropic documents public, revealing Claude Mythos — a model that outperforms Opus 4.6 across cybersecurity, coding, and reasoning benchmarks. Internal materials describe it as occupying a new tier above Opus. Anthropic says it's extremely compute-intensive and is working on efficiency before general release.

RELEASE2026-04-02

Cursor ditches the code editor paradigm — Cursor 3 is an agent orchestration platform where you dispatch parallel AI workers, not write code

7
Read more

Developed under the code name Glass, Cursor 3 replaces the traditional IDE with an agent-orchestration layer. Developers describe tasks in natural language, spin up parallel agents, and review outcomes before committing. New features include an Agents Window, Design Mode, native best-of-N model comparison, and cloud-local agent handoff. The move comes as Claude Code holds 54% of the AI coding market.

HOT TAKE2026-03-29

Sora burned $1M per day with under 500K users — OpenAI pulled the plug, killed a $1B Disney partnership, and redirected compute to coding and enterprise

7
Read more

OpenAI announced Sora's two-stage shutdown: app closes April 26, API follows in September. After peaking at ~1M users post-launch, retention collapsed below 500K while daily compute costs hit $1M. Disney's $1B partnership died with it — they found out less than an hour before the public. The pivot signals OpenAI doubling down on Codex and enterprise agents over creative tools.

NEWS2026-03-31

Anthropic, OpenAI, Google, Microsoft, and Amazon jointly founded the AAIF — MCP is now a Linux Foundation project with 146 members

6
Read more

The Linux Foundation formed the Agentic AI Foundation (AAIF) with three anchor projects: Anthropic's MCP, Block's goose agent framework, and OpenAI's AGENTS.md. Platinum members include AWS, Anthropic, Block, Bloomberg, Cloudflare, Google, Microsoft, and OpenAI. The foundation has grown to 146 members including American Express, JPMorgan Chase, and Red Hat.

NEWS2026-04-03

Anthropic paid $400M in stock for a 10-person ex-Genentech team — the bet is that domain-specific biotech AI will be worth enterprise prices

5
Read more

Anthropic acquired Coefficient Bio, a stealth startup with fewer than 10 people — nearly all former Genentech computational biology researchers. The team built a platform for AI-driven drug R&D planning, clinical regulatory strategy, and drug candidate discovery. They join Anthropic's Healthcare Life Sciences unit, signaling a push into specialized, high-margin verticals beyond general-purpose AI.