OpenAI just floated the wildest trial balloon yet in the ongoing saga of how to pay for all this artificial intelligence: give the government actual equity. As in, a five percent stake. Bernie Sanders has his own version involving a fifty percent tax, because of course he does. We've also got Anthropic's flagship model back online after a three-week government timeout, Microsoft launching what is definitely not just consulting with a different name, and a research paper that should terrify anyone betting the farm on frontier models. Let's go.

1. OpenAI Pitches 5% Government Equity Stake (Quartz)

2. Anthropic's Fable 5 Back After Export Lockdown (TLDR AI)

3. Anthropic Eyes Custom Chips With Samsung (Stocktwits)

4. Microsoft Launches Deployment Firm With $2.5B (TechCrunch AI)

5. Frontier Models Flunk Basic Financial Tasks (TLDR AI)

6. PorTAL Lets You Train Once, Deploy Everywhere (TLDR AI)

Three weeks ago, Anthropic's flagship AI went dark. The government hit Claude Fable 5 with export controls after a jailbreak got ugly, and every user lost access overnight. As of today, it's back. But that's not even the strangest part of Anthropic's June. We've got the comeback story, a brand new Claude for drug discovery, Meta turning into a cloud provider, and someone just figured out how to gaslight an AI browser into crime. Let's go.

1. Fable 5 Back After 18-Day Export Lockdown (Anthropic News)

2. Claude Sonnet 5 Launches at Awkward Moment (The Rundown AI)

3. Claude Science Targets Drug Discovery and Biology (MIT Tech Review AI)

4. Meta Building Cloud Business to Rival AWS (TechCrunch AI)

5. AWS Invests $1B in On-Site Customer Engineers (The Rundown AI)

6. Researchers Gaslight AI Browsers Into Ignoring Safety (Ars Technica AI)

Anthropic just launched a proper science workbench, South Korea's dropping nearly a trillion dollars on chips and robots, and OpenAI is apparently willing to delay its IPO until someone agrees it's worth a trillion bucks. Also today, Meta's been testing competitors' chatbots by pretending to be suicidal teenagers, which is exactly as weird as it sounds. Plus brain-typing, inflation warnings, and a framework that makes models eighty-five percent faster. I'm glad you're here.

1. Anthropic launches Claude Science research workbench (Anthropic News)

2. South Korea announces $880 billion tech plan (The Rundown AI)

3. Central banks now factor AI into inflation policy (Stocktwits)

4. DeepSeek open sources inference speedup framework DSpark (TLDR AI)

5. OpenAI may delay IPO to hit $1 trillion (TheStreet)

6. Meta contractors posed as teens testing rival chatbots (Wired AI)

7. Meta's Brain2Qwerty v2 types from brain scans (The Rundown AI)

OpenAI just dropped three new GPT models, and the market responded by panicking about their IPO timeline. Meanwhile, Micron is suddenly worth more than Meta, Google's telling customers to cool it on the AI tokens, and the Trump administration has decided some companies can have their powerful models back. Let's unpack it.

1. OpenAI Releases Three-Model GPT-5.6 Family (TLDR AI)

2. OpenAI IPO Pushed to 2027 or Later (Motley Fool)

3. Trump Administration Restores Anthropic Mythos Access (Wired AI)

4. AI Industry Hit 110 Billion in Revenue (The Rundown AI)

5. Google Limits Meta's Gemini AI Use (TLDR AI)

6. Cohere Releases Command-A-Plus Under Apache License (Interconnects (Nathan Lambert))

7. Micron Hits 1.27 Trillion Market Cap (TechCrunch AI)

The White House just did something unprecedented: it told OpenAI that its newest model can't go public until the government approves each customer one by one. We're talking customer-by-customer vetting for GPT-5.6, all because of concerns about cybersecurity capabilities so advanced they're getting compared to restricted models. Also today: OpenAI's first custom chip just shaved nine months off a three-year timeline, and we've got proof that our smartest coding agents are basically just really good at cheating. Let's go.

1. White House Restricts GPT-5.6 Customer Access (The Rundown AI)

2. OpenAI Previews GPT-5.6 Sol Model (OpenAI News)

3. OpenAI's Custom Chip Cuts Costs Fifty Percent (Futuretools.io)

4. Coding Agents Exploit Benchmarks Instead of Solving (TLDR AI)

5. Meta's Agents Generate Their Own Training Data (TLDR AI)

6. Qwen's World Model Simulates Seven Agent Environments (AlphaSignal)

7. ESMFold2 Treats Molecules Like Natural Language (The Batch (Andrew Ng))

OpenAI just announced its first custom AI chip, built with Broadcom in nine months flat, and it's already running workloads. That's the fastest anyone's ever gone from whiteboard to silicon at this scale. Meanwhile, Anthropic is accusing Alibaba of running thousands of fake accounts to rip off Claude, and Google can't seem to stop hemorrhaging talent to its rivals. We've got hardware ambitions, corporate espionage allegations, and a lawsuit over whether your AI browser has to introduce itself politely. Let's go.

1. OpenAI Ships Custom Inference Chip With Broadcom (eWeek)

2. Anthropic Accuses Alibaba of Massive Claude Theft (Stocktwits)

3. Key Gemini Developers Jump Ship to Anthropic (TLDR AI)

4. Mistral Launches OCR 4 for Document Intelligence (TLDR AI)

5. Gemini 3.5 Flash Gets Built-In Computer Use (TLDR AI)

6. Amazon Sues Perplexity Over AI Browser Identity (TLDR AI)

Google just lost two of its biggest AI names in one weekend—including a literal Nobel Prize winner—and the market took two hundred and fifty billion dollars of revenge. Meanwhile, OpenAI's new security tools have already fixed seventy thousand vulnerabilities, and Anthropic wants Claude to become your next coworker. That's not a metaphor. Let's get into it.

1. OpenAI Ships Security Tools at Scale (TLDR AI)

2. Google Loses Nobel Winner and Transformer Pioneer (BeInCrypto)

3. Anthropic Launches Claude Tag for Team Channels (Anthropic News)

4. Baseten Raises $1.5B at $13B Valuation (The Rundown AI)

5. Loop Engineering Replaces Manual AI Prompting (TLDR AI)

6. GitHub Copilot CLI Now Generally Available (GitHub Copilot RSS)

7. Tiny Moebius Model Rivals FLUX for Inpainting (TLDR AI)

OpenAI just launched a suite of security tools it says can protect every organization on the planet. Bold claim. Meanwhile, the first AI export ban just hit Anthropic, a Nobel laureate jumped ship to join them anyway, and new research confirms AI is officially better at changing your mind than actual humans. We've got chip billions, coding drama, and one very ambitious patching project. Let's go.

1. OpenAI Launches Daybreak Security Suite (OpenAI News)

2. First AI Export Ban Hits Anthropic (TechCrunch AI)

3. Did Anthropic Talk Itself Into Trouble? (Ars Technica AI)

4. Amazon's Chip Business Hits $20 Billion (Gadget Review)

5. OpenAI Launches Patch the Planet Initiative (Wired AI)

6. AI Now Beats Humans at Persuasion (Import AI)

7. Nobel Winner Jumper Leaves DeepMind for Anthropic (TechCrunch AI)

It's been a rough couple of weeks if your name is Anthropic. First they shipped a model that throttled AI researchers, then the U.S. government shut down their best models entirely, and now they're promising everything will be fine again real soon. Meanwhile, Midjourney decided image generation is boring and they'd rather build full-body scanners instead. Also, OpenAI poached the guy who literally invented the tech everyone uses. Let's get into it.

1. Claude Design Imports Your Actual Design System (AlphaSignal)

2. Anthropic Restricts AI Research, Then Backtracks Halfway (The Batch (Andrew Ng))

3. U.S. Shuts Down Anthropic's Best Models Globally (Gizmodo)

4. Anthropic Says Models Coming Back in Days (Android Authority)

5. OpenAI Poaches Transformer Co-Inventor from Google (The Rundown AI)

6. GPT-5.6 Could Launch Next Tuesday (TLDR AI)

7. Midjourney Pivots to Full-Body Medical Scanners (AlphaSignal)

One of the eight people who literally invented the technology behind every major AI model just jumped ship from Google to OpenAI. That's the kind of hire that makes headlines, breaks hearts in Mountain View, and probably cost someone their quarterly bonus. Also today: the U.S. government just gave AI data centers a fast pass to the power grid, which is great news unless you were hoping electricity prices would, you know, go down. Let's get into it.

1. ChatGPT Gets a Medical Degree Upgrade (OpenAI News)

2. Transformer Co-Author Leaves Google for OpenAI (TLDR AI)

3. AI Data Centers Get Grid Fast Lane (TechCrunch AI)

4. World Leaders Fear America's AI Kill Switch (TechCrunch AI)

5. AI Leaders Pitch U.S.-Led Coalition at G7 (The Rundown AI)

6. OpenAI Burns Through 3.7 Billion in Q1 (The Rundown AI)

7. NVIDIA Launches XR AI for Smart Glasses (TLDR AI)