Latest News

DeepSeek V4 drops two open-weight models with 1M-context by default, CSA+HCA hybrid attention, and V4-Pro priced at roughly 1/7 Opus 4.7's output cost.

April 26, 20262 min read

IndustryMajor

GPT-5.5: Agentic-First Model, 82% Terminal-Bench, Safety at HIGH

OpenAI's GPT-5.5 arrives six weeks after 5.4 with a 7-point Terminal-Bench gain, doubled pricing, and cyber/bio safety classifications at HIGH.

April 26, 20262 min read

Industry

SpaceX's $60B Cursor Option: Frontier-Lab API Dependency as Weaponizable Risk

SpaceX holds a $60B call option on Cursor with a $10B breakup floor — structured as much against frontier-lab API dependency as toward acquisition.

April 24, 20262 min read

Industry

Google Reframes Vertex AI as Gemini Enterprise Agent Platform at Cloud Next

Google rebrands Vertex AI as the Gemini Enterprise Agent Platform at Cloud Next, adding 200+ models and five major consulting partnerships to close the enterprise adoption gap.

April 23, 20262 min read

Industry

OpenAI Offers Free ChatGPT to US Clinicians, Releases HealthBench

OpenAI makes ChatGPT free for verified US medical professionals and releases HealthBench Professional — an open benchmark that GPT-5.4 outperforms on against physicians.

April 23, 20262 min read

Industry

OpenAI Launches ChatGPT Workspace Agents for Enterprise Plans

OpenAI launches ChatGPT Workspace Agents in research preview — cross-tool automation agents for Business, Enterprise, Edu, and Teachers plans.

April 23, 20262 min read

Industry

Anthropic's Compute Shortfall Has Become a Trust Crisis

Dario Amodei's conservative 2024 capex call is now triggering opaque quota changes, harness bans, and tokenizer inflation — handing OpenAI a sustained PR windfall.

April 23, 20262 min read

Industry

OpenAI Launches GPT-5.1 with Native Agent Infrastructure

OpenAI released GPT-5.1 with built-in agent orchestration, persistent memory, and native tool invocation — positioning against Anthropic and Google in the agentic AI infrastructure race.

April 14, 20263 min read

Technology

6 articles

View all in Technology

TechnologyNotable

Kimi K2.6 Becomes Open-Source #1 with 300-Agent Swarms

Moonshot AI's Kimi K2.6 leads the open-source index with 300 concurrent sub-agents, 4,000 tool calls, and a 12-hour autonomous coding marathon.

April 26, 20262 min read

TechnologySignificant

Google Ships Deep Research Max with MCP and $4.80 Per-Report Pricing

Google Deep Research Max costs $4.80/report and uses MCP to connect to private data stores. Independent 7-task testing shows the cheaper tier wins 5 of 7.

April 26, 20262 min read

Technology

GPT Image 2 Wins 93% of Blind Tests — Reasoning Joined the Visual Stack

GPT Image 2 claims a 26-point lead in Image Arena blind tests — unprecedented for the category — by wiring a reasoning loop before every pixel render.

April 25, 20262 min read

Technology

DeepSeek V4-Pro: 10× KV Cache Efficiency at Open-Source Scale

DeepSeek V4-Pro launches with 1.6T parameters, 1M context, and 10× KV cache reduction over V3.2 — multiplying inference concurrency roughly 10× on the same hardware.

April 24, 20262 min read

Technology

GPT-5.5 Reframes AI Progress as Intelligence Per Token

GPT-5.5 scores 2.5× better intelligence-per-token than 5.4, surpasses the human baseline on OS World, and expands Codex into a full desktop agent.

April 24, 20262 min read

Technology

Qwen3.6-27B Surpasses a 397B Model on Coding Benchmarks

Alibaba's Apache 2.0 27B model outperforms Qwen3.5-397B-A17B on all major coding tasks and runs locally on 18 GB RAM — 'bye bye subscription era' claims are spreading.

April 23, 20262 min read

Strategy

2 articles

View all in Strategy

Strategy

Anthropic's Project Deal: Agents Closed 186 Trades — Humans Couldn't Tell the Difference

Anthropic ran a live two-sided agent marketplace with 69 employees: 186 deals, $4,000+ volume — and model quality (Opus vs Haiku) was invisible to human participants throughout.

April 25, 20262 min read

Strategy

Token Theater: How Enterprises Are Measuring AI Adoption Backwards

Matt Shumer's viral thread documents a new enterprise failure mode: promotions and firings based on tokens consumed rather than output shipped — and predicts an 18-month ROI reversal.

April 23, 20262 min read

Regulation

1 article

View all in Regulation

Regulation

EU AI Board Publishes First Implementation Guidelines

The European AI Board released its first formal implementation guidelines, clarifying risk classification criteria and technical documentation requirements ahead of the August 2026 enforcement date.

April 13, 20262 min read

Research

3 articles

View all in Research

Research

Virginia Tech Preprint Challenges Skill-MD Paradigm with Model-Native Training

A Virginia Tech preprint shows model-native skills extracted via sparse autoencoders outperform human-defined skill files in SFT — and produce 41% gains on math via activation-space data selection.

April 25, 20262 min read

Research

MILKYWAY Shows Agent Scaffolding Can Outperform Fine-Tuning

A new paper freezes GPT-5.4's weights and puts all learning in an editable text harness, hitting 61% on prediction benchmarks where the base model scores 44%.

April 23, 20262 min read

Research

Meta Open-Sources Llama 4 with Native Context Engineering

Meta released Llama 4 under its updated open-source license, featuring built-in context engineering primitives and a 2M token context window — a significant milestone for the open-source LLM ecosystem.

April 11, 20263 min read

Tools

4 articles

View all in Tools

Tools

Matt Pocock's Counter-Thesis: The Codebase Is the Agent's Ceiling

Matt Pocock's two-hour AI Engineer workshop argues 30-year-old software fundamentals matter more under AI, not less — and outlines a complete methodology to prove it.

April 25, 20262 min read

Tools

Claude Code Regression: Three Harness Issues, One Public Post-Mortem

Anthropic published a post-mortem on three sequential Claude Code harness changes from March–April that degraded output quality, fixed in v2.1.116+.

April 24, 20262 min read

Tools

ml-intern: HuggingFace Releases a Full-Loop Autonomous Post-Training Agent

ml-intern reads arXiv, cleans datasets, runs SFT/GRPO, diagnoses failures, and iterates — pushing GPQA from 10% to 32% in under 10 hours for roughly $1 of compute.

April 23, 20262 min read

Tools

Anthropic Adds Persistent Memory to Claude Enterprise

Anthropic's Claude Enterprise tier now includes cross-session persistent memory, bringing it into direct competition with OpenAI's newly announced GPT-5.1 memory features.

April 12, 20262 min read

AI Intelligence Newsletter

Curated AI insights — sent when there's something worth your inbox.