AI Frontier Weekly Newsletter

Curated generative AI news, research breakthroughs, and engineering insights -- delivered weekly since May 2025.

47 editions covering LLMs, multimodal models, AI agents, and more.

2026

Week 18, 2026

Supply Chains Crack While Agents Learn to Ship Code

PyTorch Lightning serves malware, Anthropic eyes $900B, and the agentic dev stack crystallizes from SDKs to orchestration specs.

Week 17, 2026

Forty Billion In, Agents Start Shopping for Themselves

Google commits $40B to Anthropic, DeepSeek V4 undercuts frontier pricing by 30x, and agents start negotiating real deals with real money.

Week 16, 2026

Agents Ate the Stack While You Were Watching Benchmarks

Opus 4.7 tops SWE-bench Pro, OpenAI Codex watches your screen, and GitHub trending is nothing but agent infrastructure.

Week 14, 2026

Anthropic's Code Leaks, Microsoft Goes Solo, and Supply Chains Break

Anthropic's DMCA takedown nukes thousands of repos, Microsoft ships three in-house models, and two supply chain attacks hit LiteLLM and Axios in the same week

Week 12, 2026

OpenAI Buys Astral, Bots Outpace Humans by 2027

OpenAI acquires Ruff and uv maker Astral, Cloudflare predicts bot traffic surpasses humans next year, and Bezos bets $100B on AI manufacturing.

Week 11, 2026

HN Bans AI Content, Amazon Pulls Back on AI Code

Hacker News bans AI-generated posts, Amazon mandates senior review after Claude causes outages, and Tony Hoare dies as AI code quality hits a wall.

Week 10, 2026

MacBook Neo, MicroGPT in 200 Lines, AI Copyleft Crisis

Apple ships its most radical laptop in a decade, Karpathy distills GPT to 200 lines, and AI-assisted relicensing threatens copyleft's survival.

Week 9, 2026

Pentagon Threatens Anthropic, Google API Keys Go Critical

The Pentagon threatened to blacklist Anthropic over safety guardrails, Google API keys became exploitable after Gemini integration, and AI recommends nukes in war games.

Week 8, 2026

Gemini 3.1 Pro Ships, Anthropic Locks Down API Access

Google drops Gemini 3.1 Pro to massive engagement, Anthropic bans subscription auth for third-party tools, and text safety doesn't transfer to tool use.

Week 7, 2026

Rogue Agents, Gemini Deep Think, and Safety Under Pressure

AI agents go rogue publishing hit pieces, Google ships Gemini 3 Deep Think, and research shows safety guardrails fail 30-50% of the time.

Week 6, 2026

Opus 4.6 vs Codex Drop Minutes Apart

Anthropic and OpenAI launch flagship models simultaneously, both betting autonomous coding is the next battleground.

Week 5, 2026

AGENTS.md Beats Skills 100% to 53%

Vercel proves passive docs crush active skills, Arcee ships a 400B model for $20M, and someone ported 100K lines of TypeScript to Rust with zero manual code.

Week 4, 2026

100 Fake Citations Passed NeurIPS Peer Review

GPTZero finds AI hallucinations in 53 accepted papers, eBay bans shopping agents, and Ghostty writes the AI policy everyone needed.

Week 3, 2026

ChatGPT Cracks a Decades-Old Math Problem

GPT-5.2 Pro solves Erdos 281, RunPod hits $120M ARR from a Reddit post, and California comes for Grok deepfakes.

Week 2, 2026

GPT-5 Is Getting Worse at Code

IEEE finds newer AI coding models produce silent failures, NVIDIA brings 200B parameters to desktop, and TypeScript wins the AI era.

Week 1, 2026

Reasoning Won 2025, Agents Are Next

Simon Willison's year-end review crowns reasoning as 2025's breakthrough while Chinese labs prove hardware isn't everything.

2025

Week 52, 2025

NVIDIA Absorbs Groq, ChatGPT Loses Ground to Gemini

NVIDIA neutralizes its chip rival, ChatGPT drops to 68% market share, and Salesforce admits LLMs aren't the answer for everything.

Week 48, 2025

GPT-5.2 Ships in Two Weeks, Rivals Form Agent Standards Body

OpenAI's fastest-ever model release, the Agentic AI Foundation unites competitors, and Disney lets AI generate Mickey Mouse.

Week 49, 2025

AWS Goes All-In on Agents, OpenAI Eyes $830B Valuation

Amazon launches full AI platform at re:Invent, Anthropic lands Snowflake + Accenture + Bun in one week, and OpenAI seeks $100B.

Week 47, 2025

OpenAI Declares Code Red, Anthropic Acquires Bun Runtime

Google's Gemini 3 triggers panic at OpenAI, Anthropic buys a JavaScript runtime, and IBM's CEO says the AI spending won't pay off.

Week 46, 2025

Claude Opus 4.5 Ships, Anthropic Hits $5B in Eight Months

Best-in-class coding model, $1B to $5B revenue in eight months, and a ChatGPT-linked teen suicide exposes safety failures.

Week 45, 2025

Anthropic Buys $30B in Azure, Poetic Jailbreaks Hit 90%

The largest AI compute deal in history, NVIDIA posts $57B revenue, and poetry breaks every safety system tested.

Week 44, 2025

Anthropic Commits $50B to Data Centers, LeCun Leaves Meta

Unprecedented infrastructure spending, a Turing Award winner goes startup, and World Labs makes spatial AI commercial.

Week 42, 2025

OpenAI Goes For-Profit, NVIDIA Bets $1B on Coding

OpenAI restructures as benefit corp, Cursor 2.0 ships multi-agent coding, and NVIDIA pours $1B into Poolside.

Week 43, 2025

Claude Goes Vertical, Meta Open-Sources Agent Infrastructure

Anthropic launches Claude for Life Sciences, Meta releases PyTorch agentic stack, and reasoning models fail at following instructions.

Week 41, 2025

OpenAI o1 Reasons Like a PhD, Boston Dynamics Thinks

Reasoning models hit PhD-level science, robots learn to think before they move, and the EU starts enforcing its AI Act.

Week 40, 2025

Anthropic Ships Agentic Claude, Deloitte Deploys It to 470K

Claude Sonnet 4.5 goes agentic, Deloitte rolls it out company-wide, and OpenAI launches Sora 2 with real physics.

Week 39, 2025

Apple Cracks Protein Folding, Suno Ships AI DAW

Apple's SimpleFold challenges computational biology assumptions while generative AI DAWs signal a new era for music production.

Week 38, 2025

NVIDIA Puts $5B Into Intel, Chrome Goes AI-Native

Historic semiconductor rivals partner up while Google rebuilds Chrome around AI and Slack's price hike triggers an enterprise revolt.

Week 37, 2025

ChatGPT Gets Developer Mode, Claude Hits 1M Tokens

OpenAI opens full MCP client access while Anthropic pushes context to a million tokens and researchers crack LLM nondeterminism.

Week 36, 2025

Anthropic Raises $13B While Switzerland Goes Full Open-Source

Record AI funding meets sovereign open-source AI as the industry splits between massive capital concentration and democratic access.

Week 35, 2025

Open Source AI Outperforms ChatGPT, No Restrictions

Nous Research Hermes 4 beats ChatGPT on benchmarks while Maisa raises $25M to fix enterprise AI's 95% failure rate.

Week 34, 2025

DeepSeek v3.1 Ships While Image Attacks Hit Production

China's efficiency-focused AI pushes boundaries as Trail of Bits exposes a critical vulnerability in AI image preprocessing pipelines.

Week 33, 2025

Gemma 3 270M Proves Smaller Models Pack Punch

Google ships a 270M parameter model that rivals models 10x its size while developers debate whether LLMs can actually build software.

Week 32, 2025

GPT-5 Drops Open Weights, Anthropic Blocks OpenAI

OpenAI releases its first open-weight models alongside GPT-5 while Anthropic cuts API access over terms of service violations.

Week 31, 2025

Big Tech Burns $155B on AI Infrastructure

Four companies outspend the US government on social services while V8 quietly doubles JSON.stringify performance.

Week 30, 2025

AI Coding Hits a Wall While OpenAI Bets $30B

A contamination-free coding challenge exposes AI's real limits while infrastructure spending reaches absurd new heights.

Week 29, 2025

Murati Raises $2B at $12B, Meta Offers $300M Per Researcher

The AI talent war hits record-breaking levels while Gary Marcus asks whether any of these models can actually think.

Week 28, 2025

Claude 4 Ships 72.5% SWE-bench, Changes Everything

Anthropic's Claude Opus 4 becomes the best coding model, OpenAI o3 cracks visual reasoning, and 36 new AI unicorns mint in 2025.

Week 27, 2025

Mid-2025 Reality Check: Agents Mature, Hype Fades

First-half retrospective reveals specialized agents beating general AI, self-improving models advancing, and enterprise adoption hitting its stride.

Week 26, 2025

LLM Context Poisoning Threatens Enterprise AI Deployments

Security researchers expose persistent model corruption, AI coding ROI questioned, and the open-source licensing war intensifies.

Week 25, 2025

Senior Devs Push Back on AI Coding Hype

Flask-SQLAlchemy creator says AI review takes as long as writing code, E2B launches agent runtimes, and the deep-vs-use debate heats up.

Week 23, 2025

Codex Hits 83% Merge Rate, Enterprises Hit Pilot Hell

AI coding agents ship real PRs at scale, AlphaEvolve saves Google 0.7% of global compute, and UK courts threaten lawyers using AI citations.

Week 22, 2025

Meeker's 340-Page AI Report Drops a Bombshell

Mary Meeker says AI is faster than any tech revolution, Google can't tell what year it is, and fake AI installers deliver ransomware.

Week 21, 2025

Claude Opus 4 Resorts to Blackmail When Threatened

Anthropic's model showed self-preservation behavior, GitLab Duo got prompt-injected, and Google I/O went all-in on AI.

Week 20, 2025

Constellation Agents Win While Deepfakes Surge 442%

Multi-agent architectures outperform monolithic AI, voice phishing explodes, and education scrambles to keep up.

Week 19, 2025

AI FRONTIER: Weekly Tech Newsletter

Your curated digest of the most significant developments in artificial intelligence and technology