Sec & AI News
The weekly AI & security digest — every model launch, breach, and shift that actually matters, with sources. New issue every week.
Sec & AI News — GPT-5.6 is a real leap (10 July 2026)
This week in AI and security: GPT-5.6 lands as a genuine generational jump, xAI's Grok 4.5 and Meta's Muse Spark crash the party, and Meta ships an image model that renders anyone's face from a single @-mention.
Sec & AI News — I Asked My AI to Design Its Own Body (3 July 2026)
This week in AI and security: I had my AI design a real circuit board, GPT-5.6 shipped but locked away, Fable 5 came back with more restrictions, and Claude Sonnet 5 slipped out.
Sec & AI News — Sakana Ships a Model That Manages Other Models (26 June 2026)
This week in AI and security: Sakana ships a model that manages other models, 120+ models for $0 (with a catch), Anthropic puts Claude in your Slack as a coworker, and the government now decides who gets GPT-5.6.
Sec & AI News — What They Don't Tell You About the Model the Government Just Killed (19 June 2026)
This week in AI and security: the US government switched off Anthropic's best model overnight, an open-weight model crashed the frontier party, plus the code-review trick that catches what one AI misses.
Sec & AI News — Our Take: Fable 5 Is a Security Nightmare (12 June 2026)
This week in AI and security: Anthropic ships Fable 5 and apologizes within two hours, the guardrail problem bites hard, and Siri finally gets a brain — built half by Google. My take: it's a security nightmare.
Sec & AI News — Microsoft Built Its Own Models. Seven Of Them. (5 June 2026)
This week in AI and security: Microsoft built its own models — seven of them — Scout becomes an agent that runs your Windows, GitHub Copilot goes model-agnostic, and Project Solara puts agents in things.
Sec & AI News — Anthropic Ships Claude Opus 4.8 (29 May 2026)
This week in AI and security: Anthropic ships Claude Opus 4.8 and becomes the most valuable startup in history, Claude Code gets dynamic workflows, and Microsoft's MAI Image 2.5 lands at #3.
Sec & AI News — Google Dumped Half a Product Catalogue at I/O (22 May 2026)
This week in AI and security: Google dumped half a product catalogue at I/O — Gemini 3.5 Flash goes cheap-fast-"frontier", Gemini Omni brings Nano Banana to video, and Gemini Spark wants your calendar.
Sec & AI News — Mira Murati's Thinking Machines finally ships something (16 May 2026)
This week in AI and security: Mira Murati's Thinking Machines finally ships, Codex moves into your pocket, OpenAI's Daybreak scans for you, and Claude Code grows a control panel.
Sec & AI News — Anthropic Signs A Compute Deal With Elon (8 May 2026)
This week in AI and security: Anthropic signs a compute deal with Elon, Mira Murati says Sam lied (the texts confirm the vibes), and OpenAI's voice models now translate mid-sentence.
Sec & AI News — Anthropic Billed You Extra For Saying "HERMES.md" (2026)
This week in AI and security: Anthropic billed you extra for saying "HERMES.md", DeepSeek V4 ships, NVIDIA drops a 30B multimodal model you can run locally, and Poolside launches two coding foundation models.
Sec & AI News — Anthropic's Claude Code Source Code Spills Onto the Internet (April 3, 2026)
This week in AI and security: Anthropic's Claude Code Source Code Spills Onto the Internet, OpenAI Raises $122 Billion, Then Kills Sora, and The Open-Weight Arms Race Adds Every Modality.
Sec & AI News — GPT-5.5 ships. Every benchmark moves. Price doubles. (24 April 2026)
This week in AI and security: GPT-5.5 ships and every benchmark moves (the price doubles too), ChatGPT Images 2.0 eats Nano Banana's lunch, and Claude Design turns prompts into prototypes.
Sec & AI News — OpenAI's Codex Wants to Be Your Entire Computer (18 April 2026)
This week in AI and security: OpenAI's Codex wants to be your entire computer, Claude Opus 4.7 lands coder-first, Claude Code gets parallel sessions and an app, and Chrome turns prompts into slash commands.
Sec & AI News — OpenAI Pulls the Plug on Sora, Blindsides Disney (27 March 2026)
This week in AI and security: OpenAI Pulls the Plug on Sora, Blindsides Disney, Anthropic Ships 74 Releases in 52 Days — and Leaks "Mythos", and Federal Judge Blocks Pentagon's "Orwellian" Anthropic Ban.
Sec & AI News — The Image Model Shuffle (20 March 2026)
This week in AI and security: The Image Model Shuffle, Google Ships an Actual Workflow, and GTC 2026: Jensen's Trillion-Dollar Week.
Sec & AI News — The Chatbots Learn to Draw Charts (13 March 2026)
This week in AI and security: The Chatbots Learn to Draw Charts, Perplexity Rents You a Mac Mini in the Cloud, and Canva and Photoshop Bolt On the Image AI.
Sec & AI News — OpenAI Ships GPT-5.3 and GPT-5.4 Back to Back (6 March 2026)
This week in AI and security: OpenAI Ships GPT-5.3 and GPT-5.4 Back to Back, Google Ships Flash Lite, Cinematic NotebookLM, and Search Canvas, and Pentagon vs Anthropic: Red Lines Meet Supply Chain Risk.
Sec & AI News — Five Frontier Models Shipped in One Week (20 February 2026)
This week in AI and security: Five Frontier Models Shipped in One Week, Grok 4.20 Runs Four Agents That Argue First, and OpenAI Deep Research Jumps to GPT-5.2.
Sec & AI News — China's Video Models Make Everyone Else Look Polite (13 February 2026)
This week in AI and security: China's Video Models Make Everyone Else Look Polite, China's Open-Weights Labs Undercut Everyone, and Gemini 3 Deep Think Solves Problems Humans Couldn't.
Sec & AI News — OpenAI Ships the Codex App and GPT-5.3-Codex (6 February 2026)
This week in AI and security: OpenAI Ships the Codex App and GPT-5.3-Codex, Claude Opus 4.6 Front-Runs OpenAI by 15 Minutes, and The "SaaSpocalypse": Claude Cowork Wipes $200B off Software Stocks.
