DAILY INTELLIGENCE BRIEF

The Vault -- AI Edition

Monday, July 13, 2026
15 stories
13,754 words
70 min read

Business

5
01

Google and Microsoft Back a Rival Agent Protocol to Counter Anthropic and OpenAI in the Enterprise

The two companies that spent a decade fighting each other over cloud contracts have found a common enemy in the enterprise.

5 minRead →
02

Musk and Altman Turn Apple's OpenAI Lawsuit Into a Public Feud on X

Apple's lawyers filed the complaint on Friday. By Saturday morning, the two most combustible personalities in AI had turned it into a brawl.

4 minRead →
03

OpenAI's Deployment Company Acquires Northslope to Build Out Forward-Deployed Engineering

OpenAI's enterprise arm is buying its way into the consulting business one embedded engineering shop at a time.

5 minRead →
04

Cloudflare Opens the Waitlist for an x402 Gateway That Lets AI Agents Pay for the Web

For nearly three decades, one corner of the web's plumbing has sat idle: HTTP status code 402, 'Payment Required'.

4 minRead →
05

Venice Raises $65 Million to Build Private, Surveillance-Free Access to AI Models

For two years, Venice grew into a $1 billion company almost entirely off its own revenue, without a dollar of outside capital.

5 minRead →

Llms Genai

4
06

Mistral's Leanstral 1.5 Brings Formal Mathematical Proof to Code Verification

A widely used Rust library had a bug hiding in plain sight — until an AI wrote a mathematical proof that caught it.

4 minRead →
07

A GPT Image 2 and Runway Aleph 2.0 Workflow Turns Video Editing Into a One-Frame Job

A filmmaker shoots a chase scene, then decides the hero's jacket is the wrong color. Two years ago that was a reshoot.

5 minRead →
08

Goldman Sachs Tells Clients Which Chinese AI Models to Use, Marking Their Enterprise Normalization

When one of Wall Street's most conservative institutions hands clients a shopping list of Chinese AI models, the era of curiosity is over.

4 minRead →
09

Anthropic Triples Project Glasswing to 150 Organizations Across 15 Countries

Three weeks after the Five Eyes alliance said AI-fueled cyberattacks were months, not years, away, Anthropic scaled its answer.

5 minRead →

Research

3
10

NVIDIA Launches an Open Healthcare-Robotics Platform With Open-H and Cosmos-H

For a decade, medical AI has mostly been a spectator: it reads scans and flags nodules. NVIDIA wants it to pick up the instruments.

5 minRead →
11

Proteina-Complexa, a Protein-Design Reasoning Model, Validates a Million AI-Designed Binders

For years, the standard rebuttal to any AI-for-biology result was the same: it looks great in silico, but does it hold up in the lab?

5 minRead →
12

A UK Government Lab Found Universal Jailbreaks in GPT-5.6's Cyber Safeguards

Days after OpenAI billed GPT-5.6 Sol as its most secure model ever, British government researchers broke it within hours.

5 minRead →

Policy

3
13

The ECB Warns AI Could Make Inflation More Volatile

Central bankers spent the 2010s worrying inflation was too low and the 2020s fighting to bring it down. AI may make it jumpy in both directions.

5 minRead →
14

The AI Labor Reckoning: Majorities Back a Wealth Fund as Tech Layoffs Accelerate

A supermajority of Americans now wants the government to seize a share of the equity of the largest AI companies.

5 minRead →
15

Data-Center Backlash Grows as AI's Physical Footprint Meets Local Resistance

For two years, the AI buildout answered every obstacle with money. Then it ran into neighbors who were not for sale.

4 minRead →