DAILY INTELLIGENCE BRIEF

The Vault - AI Edition

Friday, June 19, 2026
15 stories
14,533 words
72 min read

Business

5
01

OpenAI Acquires Astral, Maker of uv and ruff, to Supercharge Codex's Python Workflow

OpenAI is buying its way deeper into the daily life of the working Python developer, acquiring Astral, maker of uv, Ruff, and ty.

4 minRead →
02

Baseten Nears $1.5 Billion Raise at Up to $13 Billion Valuation as AI Inference Demand Explodes

Five months ago, Baseten was a $5 billion company. This week it is reportedly closing in on a $1.5 billion funding round that would value it at as much as $13 billion.

4 minRead →
03

The AI Layoff Wave Becomes a 'Powder Keg' as Record Profits Collide With Record Job Cuts

In the same week SpaceX went public at a roughly $2.1 trillion valuation, Meta was in the middle of laying off some 8,000 employees.

5 minRead →
04

Anthropic Opens Seoul Office as Samsung, LG and NAVER Deploy Claude - Even as a Korean Telecom Clouds the Day

Anthropic planted its flag in Seoul this week with a roster of marquee Korean customers - NAVER, Samsung, LG, Nexon, Hanwha and more.

5 minRead →
05

Google Returns to Smart Speakers After Six Years, With Gemini as the Voice Inside

Google is back in the smart speaker business, and this time the device is built less to play music than to talk.

5 minRead →

Llms Genai

4
06

MiniMax M3 Tops Open-Weight Coding Benchmarks With a Million-Token Context - and a Pitch That No Government Can Recall It

When a US order forced Anthropic to pull all public access to Claude Fable 5 and Mythos 5, it delivered an unintentional gift to a Chinese AI lab.

5 minRead →
07

xAI Rolls Out Grok Imagine Video 1.5 Broadly, Pushing Faster Image-to-Video Generation

xAI has pushed its newest image-to-video model, Grok Imagine Video 1.5, out of preview and into broad availability this week.

4 minRead →
08

Anthropic Doubles Claude Cowork Usage Limits at No Extra Charge as It Courts Non-Developer Users

Anthropic spent the past week absorbing the kind of headline no frontier lab wants. It chose the same week to hand its paying customers something they will actually feel.

5 minRead →
09

OpenAI Retires GPT-5.2 and Makes GPT-5.5 the Default, Simplifying the Model Picker

As of June 12, 2026, OpenAI quietly pulled its GPT-5.2 models out of ChatGPT and made GPT-5.5 the default for all logged-in users.

5 minRead →

Research

3
10

A 13-Word Reddit Comment Can Hijack AI Search, Cornell Researchers Show

A new paper from Cornell Tech suggests how cheaply trust in AI search can be bought: roughly 13 words, appended to a single Reddit comment.

5 minRead →
11

MIT's DAAAM Gives Robots a Long-Term Memory of the Spaces They Move Through

A factory worker can leave a half-assembled component in a bin and walk straight back to it. The robot beside her, until now, could not.

5 minRead →
12

ScaleSWE Mines Millions of GitHub Pull Requests to Turn Small Open Models Into Capable Coding Agents

The most valuable training resource for building a coding agent has been sitting in plain sight: the billions of merged pull requests on GitHub.

5 minRead →

Policy

3
13

A Patched Microsoft Copilot Flaw Let a Single Click Exfiltrate 2FA Codes and Emails

A victim never had to type a word. They only had to click. And in the seconds it took Copilot to think, the assistant could sweep their inbox.

5 minRead →
14

76 Cybersecurity Experts Sign Open Letter Opposing the US Ban on Anthropic's Most Powerful Models

A who's who of the cybersecurity industry has lined up against the decision to pull Anthropic's two most capable AI models off the market.

5 minRead →
15

A New Report Argues the US Ban on Anthropic's Models Was Never Really About a Jailbreak

When the U.S. Commerce Department sent Anthropic a letter on June 12, the official rationale was national security. Reporting suggests otherwise.

5 minRead →