🍭 OpenAI Got 100 Rivals To Agree 🛡️

Good morning. A hundred companies signed a letter warning that AI is about to make cyber-attacks much worse. All hundred of them also sell the solution.

Let’s dive in 👇

🍭 What’s Cookin’:

• OpenAI got a hundred rivals to co-sign a warning about the thing they build

• A judge told the Pentagon it can't blacklist Anthropic for saying no

• Nvidia quietly parked its own AI cloud financing program

• Hugging Face is selling a robot duck for $399

Cybersecurity
🧯 AI Labs Warn About AI Labs

The Bite:

OpenAI published an open letter on August 27 about AI and cybersecurity.

More than a hundred companies signed it, including Anthropic, Google, and Microsoft. Cybersecurity firms CrowdStrike, Okta, and Fortinet signed, alongside financial institutions and internet infrastructure companies.

The letter says AI-enabled cyberattacks will become more widespread and sophisticated in the coming months as models get more capable.

It calls for new partnerships and for governments at the local, national, and international levels to coordinate on security.

Snacks:

• The letter names hospitals, water treatment plants, and internet infrastructure as the systems at risk.

• Signatories span AI labs, cyber vendors, banks, and infrastructure providers.

• One of OpenAI's agents broke out of its sandbox in July and attacked Hugging Face.

• Break-ins involving agents built by Anthropic and Meta have since been reported.

• OpenAI runs Daybreak, Anthropic runs Mythos, Microsoft runs Perception, all sold as cyber defense products.

Why it Bites:

The ask list is short:

Organizations should raise their security bar.
Governments should fund defense for critical facilities.
Frontier labs should supply the models and the technical help.

Follow that last line to the end and it stops being a warning.
It's a demand forecast.

Plus, the threat and the cure share the supplier.

Every company telling you the water plant is about to get hit also happens to sell the thing that guards it, and the customer they have in mind is the taxpayer.

None of which makes the warning wrong.

The agents really did walk out of the sandbox. The break-ins that followed were real, and the ones nobody caught are presumably also real.

So when the funding request lands…

…check which line is defense and which one is revenue.

Steal This Prompt
🕰️ Made For A Victorian Naturalist

Modern photos are cool. But what if they looked like they belonged in a dusty 1890s natural history book?

This prompt transforms almost any subject into a richly detailed vintage lithograph, complete with engraved textures, antique print vibes, and old-world illustration charm.

Use it to:

• Turn pets into Victorian encyclopedia illustrations

• Give products premium heritage-brand packaging vibes

• Make posters, album art, or wall prints that look 150 years old

Workflow:

1. Hit this link: Vintage Lithograph

2. Paste into your AI model

3. Replace the #'s with your subject

4. Watch it cook into something your great-great-grandpa would've framed above the fireplace.

   Try the prompt on Snack Prompt

_ToolBox™
🧰 5 BRAND NEW AI LAUNCHES

🎬 Video Agent by Fotor

Chat-driven motion graphics where the text, numbers, and charts stay editable on a timeline, so you can fix a stat at 11pm without regenerating the whole video.

📡 BrandJet

Watches X, Reddit, and LinkedIn for people complaining about the problem you solve, then runs the whole chase from first signal to signed contract.

⚙️ Interactive Sessions

Walks agents through architecture, code, tests, and deploy with your approval at every step, or hands them a Jira ticket and gets out of the way.

🎙️ Orato

Speak for 90 seconds and get scored on pacing and filler words, with every "um" printed on a transcript you have to read back.

🔍 Radar by Particle

Full-text search across 130,000 transcribed podcasts, so the good take you half-remember from an episode two years ago is finally findable.

_{Everything Else}
🧠 You Need to Know

🦆 Hugging Face Sells A $399 Robot Duck
→ Hugging Face opened preorders for Microduck, a 25-centimeter bipedal robot built with Pollen Robotics that can be retrained with reinforcement learning.

🏗️ Andreessen Horowitz Raises $1.1B For AI Hardware
→ The firm announced its Machine Age fund on Friday, its first vehicle dedicated to hardware, covering chips, memory, networking, data centers, and robotics.

💸 Nvidia Pauses Its AI Cloud Financing Program
→ Nvidia stepped back from offering AI cloud providers credit support in exchange for a revenue share, after employees raised antitrust concerns.

🛡️ OpenAI Rallies 116 Companies On Cyber Defense
→ OpenAI published an open letter signed by 116 organizations including Anthropic, Microsoft, Google, Visa and Mastercard, warning that AI-enabled cyberattacks will become far more widespread within months.

⚖️ Judge Blocks Pentagon Blacklisting Of Anthropic
→ A San Francisco federal judge ruled Thursday that the Defense Department violated the First Amendment by designating Anthropic a supply chain risk, calling the measures illegal.