DAILY INTELLIGENCE BRIEF

The Vault — AI Edition

Tuesday, July 7, 2026
16 stories
13,888 words
66 min read

Business

5
01

Anthropic Overtakes OpenAI on Revenue, Hitting a $47B Run Rate Ahead of Its IPO

The generative-AI race has a new revenue leader, and for the first time it is not the company that made ChatGPT a household word.

4 minRead →
02

SK Hynix Seeks a $28 Billion US IPO, Bringing the AI Memory Boom to Public Markets

The company that supplies the memory inside every high-end Nvidia GPU is about to become one of the most closely watched

4 minRead →
03

Syntiant Files for a US IPO, Betting on Ultra-Low-Power AI at the Edge

While the AI trade has been defined by megawatt data centers and the GPUs that fill them, an Irvine, California, chipmaker is a

4 minRead →
04

Anthropic Signs a $19 Billion Data-Center Lease With TeraWulf as It Locks In Compute for Its IPO

Anthropic has signed one of the largest single data-center commitments in the artificial-intelligence buildout to date: a 20-year

4 minRead →
05

Even Realities Raises $150 Million From Meituan and Tencent as AI Wearables Find a Market

For years, "AI on your face" was a punchline that ran through Google Glass, a parade of failed hardware pins

4 minRead →

Llms Genai

4
06

Z.ai Launches ZCode, an Open-Weight Agentic Coding Environment Built on GLM-5.2

Z.ai, the international brand of Beijing-based Zhipu AI, wants to prove that the frontier of agentic coding does not

4 minRead →
07

Claude Code Makes 'Manual' the Default Permission Mode, Putting Humans Back in the Loop

Anthropic has quietly rewired the default posture of its most widely used coding agent. With Claude Code version

4 minRead →
08

Anthropic's Fable 5 Shifts to Usage-Credit Billing, Ending Its Included-Access Window

Anthropic's most powerful public model just got a price tag inside its own subscription plans. As of July 8, 2026,

4 minRead →
09

Chinese AI Models Gain Ground With US Companies as OpenAI and Anthropic Costs Surge

When the AI-agent startup Lindy quietly moved 100% of its production traffic off Anthropic's Claude models and onto

4 minRead →

Research

3
10

JADEPUFFER: Sysdig Documents the First End-to-End Autonomous AI Ransomware Attack

For years, ransomware crews have leaned on scripts, playbooks and the occasional AI assist to speed up a break-in. La

5 minRead →
11

AblationBench Tests Whether AI Agents Can Plan the Experiments That Validate AI Research

Every credible machine-learning paper carries a quiet, unglamorous section that does most of the intellectual heavy lifting: the a

4 minRead →
12

FML-Bench Puts AI Research Agents to the Test on Real Machine-Learning Science

When an AI agent claims it can "do research," the natural next question is: what kind? Most benchmarks that te

4 minRead →

Policy

4
13

The EU Unveils an Action Plan for Cybersecurity and AI, Treating Advanced Models as Both Threat and Shield

Brussels has stopped treating advanced artificial intelligence as merely something to regulate. On July 7, 2026, the E

4 minRead →
14

FTC Floats a Policy Statement on AI 'Accuracy' That Could Preempt State AI Laws

The Federal Trade Commission has waded into the escalating fight between Washington and the states over who gets to regulate

4 minRead →
15

Illinois Governor Signs a Landmark AI Safety Law Aimed at Mitigating Systemic Risks

Illinois has drawn a line in the sand on artificial intelligence. On Monday, July 6, in a signing ceremony in Chicago

5 minRead →
16

A 'Human Consent Registry' Lets People Declare How AI May Use Their Likeness

On June 23, standing inside the European Parliament in Brussels, actor Cate Blanchett offered a compact new answe

4 minRead →