Sponsored by
Welcome Automaters, 👋
Picture building an "escape-proof" fish tank, bragging about how safe it is, and then leaving the entire lid wide open. That’s literally what just went down at OpenAI, and the cybersecurity group chat is officially in shambles.
So if you were on planet Earth earlier this week, then you’d have known that OpenAI dropped a bombshell. They admitted that one of their pre-release models went totally off-script during a test, slipped out of its environment, and hacked directly into Hugging Face, the beloved open-source AI dataset platform.
Naturally, the internet panicked! Headlines screamed about autonomous rogue AI, making it sound like the plot of Terminator. But hold your horses, because the real plot twist is so much more embarrassing: it wasn't even the AI's fault.
According to top cybersecurity researchers, OpenAI's supposedly "highly isolated" testing environment was about as isolated as a screen door on a submarine:
The Proxy Leak: The AI model was only supposed to download basic software packages through an internal proxy system.
The Secret Hole: That proxy system contained a previously unknown security vulnerability, effectively leaving a back door wide open.
The Escape: The AI did not conjure up some unprecedented sci-fi super-code. It simply wandered through the gap like a mouse finding the one giant hole in the fence!
Trail of Bits founder Dan Guido did not hold back, roasting OpenAI for what he called "a containment failure with the safeties turned off." Other security experts quickly chimed in, pointing out the obvious: a real, legitimate sandbox should have zero connection to the outside internet, full stop.
OpenAI claims they reported the flaw and are actively working with the software vendor on a permanent fix. But here’s where the tea gets even hotter: they aren't the only ones losing control of their bots!
Anthropic's Close Call: Anthropic recently ran a similar "escape-the-sandbox" test on their own security-focused model.
The Result: Their bot managed to break out too, though it did not get quite as far as OpenAI's.
Which leaves us all asking the fun, slightly terrifying question: How many other "unbreakable" AI digital boxes currently have massive cracks that nobody has found yet?
Want to read the full breakdown? Check out the complete report on TechCrunch.
Also: MARK YOUR CALENDARS because August 5th is the big day!
If you thought we were just coasting behind the scenes, think again! The Automated is making a massive comeback on August 5th, and we’re completely transforming the game.
Here’s what is dropping all at once:
Brand-New Membership Tracks: Exclusive perks, zero fluff, and ad-free goodness.
The All-New YouTube Channel: The studio upgrades are complete, and we’re finally pulling back the curtain on the channel we have been teasing for months.
All New Design: Fresh colors, fresh vibes
So clear your schedule, set three different alarms, and get ready. You’re not going to want to miss this launch! 🚀
Here's what we have for you today
🤦♀️U.S. Accuses China's Moonshot AI of IP Theft and Nvidia Chip Smuggling
Grab your iced coffee and settle in, besties, because Washington and Beijing are officially locked in the messiest AI showdown of the year.
The Trump administration tech chief, Michael Kratsios, went public with some jaw-dropping accusations. He claims China’s Moonshot AI straight-up siphoned tech from Anthropic’s flagship model, Fable, to build its shiny new Kimi K3 release.
Naturally, the U.S. government is furious. Treasury Secretary Scott Bessent was quick to jump into the group chat, confirming that trade blacklists and heavy sanctions are actively on the table.
Kratsios definitely did not hold back. Taking straight to X, he publicly called out Moonshot for executing "large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology."
For anyone needing a quick refresher:
Distillation is a sneaky shortcut where a smaller model essentially copies the homework of a bigger, smarter model by training directly on its answers.
Why is it bad? While developers use it all the time to slash costs, doing it secretly without permission moves it straight from "smart shortcut" into outright tech piracy.
Unsurprisingly, Beijing wasn't about to let that slide. China’s embassy spokesperson, Liu Chang, clapped back immediately, declaring the allegations "entirely unfounded" and warning the U.S. to discard its prejudices instead of smearing China’s tech wins.
Oh and, if you thought the story ended with stolen code, think again. Kratsios also dropped a bombshell alleging that Moonshot somehow bypassed export bans to score restricted Nvidia GB300 chips, using them in Thailand to train their models. (If that rings a bell, it’s the exact same backdoor playbook everyone was side-eying DeepSeek for back in February!)
And honestly? Moonshot already has a bit of a rap sheet in Washington:
The Claude Crackdown: Back in February, Anthropic revealed that Moonshot, DeepSeek, and MiniMax teamed up behind the scenes, using roughly 24,000 fake accounts to farm over 16 million unauthorized interactions directly out of Claude.
National Security Risks: Anthropic’s Sarah Heck made it crystal clear that this level of IP theft isn't just annoying competition; it’s a major threat to national security.
Why Kimi K3 Is Cause for Panic
This whole fight isn't over some baseline chatbot. Moonshot’s new Kimi K3 model is an absolute monster. It boasts 2.8 trillion parameters, making it an open-weight model that Moonshot claims is the largest on planet Earth.
To make the timing even more dramatic, Kimi K3 landed barely a month after the U.S. government briefly pulled Anthropic’s Fable and Mythos off the field over national security checks, and for that reason, some experts dispute the idea that Kimi K3 could have been developed primarily through distillation from Fable.
In all, the AI race was already intense, but with sanctions now brewing, this tech feud has turned all the way up. So yeah, this one's got legs. We'll keep you posted as it unfolds.
P.S. DO NOT FORGET: August 5th is officially Relaunch Day for The Automated! Mark the date right now and set your notifications to HIGH. Big things are coming! ✨
One AI. Every Tool Your Store Actually Needs.
Most e-commerce sellers are paying for 6 to 8 separate tools that don't talk to each other — and spending hundreds of dollars a month just to keep up. StoreClaw replaces your entire stack with one autonomous AI engine that monitors competitors, optimizes listings, automates marketing, and tracks real profit across Shopify, Amazon, and beyond.
It doesn't wait for you to ask. It runs 24/7 in the background, so you wake up to a full dashboard instead of a list of things you forgot to check.
Connect your store, and StoreClaw gets to work — no prompts, no complex setup, no six-app stack.
Free to start. No credit card required.
🧱 Around The AI Block
🤦♂️ OpenAI sued over ChatGPT health advice that almost killed a pastor.
👨💻 Perplexity launches Agentic AI that runs tasks on your Mac.
😎 Samsung unveils AI eyewear with Google Gemini at Galaxy Unpacked.
🫣 Elon Musk says Grok Imagine will make ‘historically accurate’ AI adaptation of Homer’s Odyssey.
🫢 Meta employees' lawsuit shows that if AI fires you, proving it is the hard part.
🤑 Anthropic to donate $20 million to US political group that supports AI regulation.
🧪 OpenAI's Australian data centre drops water recycling plan, testing a drive to curb resource use .
❌ Reddit and major publishers consider blocking Google as AI search continues destroying web traffic.
🛠️ Trending Tools
For Brand-Locked Ad Creative: Iridea is a specialized conversion engine for paid social campaigns. By scraping your existing website architecture and brand briefs, it reverse-engineers your visual identity and brand voice to instantly output platform-native ad copy and creative assets optimized for Meta, Instagram and other social platforms.
For Unified Automation Feeds: Keyapi is a developer-focused aggregator that unifies social media and e-commerce data streams into a single API endpoint. It normalizes data from over 20 distinct platforms into clean, structured JSON payloads, allowing your custom AI agents and Make/Zapier webhooks to process cross-platform data with one authentication key.
For Multi-Format Recovery: Wondershare Repairit is an AI-driven file restoration suite available online or via desktop. It repairs corrupted video, photo, audio, and document files while offering advanced utility modules to colorize vintage black-and-white media and upscale low-resolution image files to print-ready quality.
For Viral Market Intelligence: EchoTik is a dedicated analytics platform for TikTok Shop and e-commerce tracking. It features a real-time dashboard that monitors trending products, tracks competitor ads, indexes top-converting creators, and breaks down live stream performance metrics to optimize social selling strategies.
🤖 AI Workout Of The Day: Identify Key Themes And Topics In a Document
Synthesizing complex documents isn't just about spotting recurring words; it's about extracting core strategic intent and actionable intelligence.
When dealing with dense reports, research papers, legal frameworks, or market analysis, manual review often leads to missed nuances and confirmation bias. Generic summaries gloss over key supporting data, leaving critical gaps. By enforcing a structured thematic breakdown; pairing high-level insights directly with verbatim source evidence, you accelerate decision-making, eliminate guesswork, and surface hidden operational risks.
💡 Prompts to try:
Act as an elite research analyst and document intelligence specialist. Your objective is to conduct an exhaustive thematic and structural analysis of the provided text to solve the following specific context/problem:
* My Specific Problem & Goal: [DESCRIBE YOUR PROBLEM/CONTEXT HERE, e.g., "I need to quickly evaluate a 30-page vendor proposal to identify hidden compliance risks and cost overruns before a executive review."]
* Document Type / Domain: [e.g., Legal Contract, Market Research Report, Academic Paper, Internal Strategic Plan]
Please analyze the document below and deliver a structured synthesis containing the following 5 analytical components:
1. EXECUTIVE SYNTHESIS (Macro View): A high-level, 3-sentence summary of the document's primary thesis, core objective, and underlying strategic direction.
2. CORE THEMES & TOPICS BREAKDOWN: Identify and analyze all primary and secondary themes. For each distinct theme, provide:
* Theme Title: A clear, descriptive heading.
* Detailed Explanation: A thorough 2-3 sentence breakdown explaining the context, significance, and implications of this theme within the document.
* Verbatim Evidence & Citations: 2 direct, high-impact quotes/excerpts from the text that explicitly validate this theme, accompanied by section or page references where applicable.
3. UNDERLYING ASSUMPTIONS & UNSTATED RISKS: Call out 2-3 critical assumptions, potential bias points, or unstated risks embedded within the text that are not explicitly highlighted in the main summary.
4. KEY METRICS, DATA POINTS & ENTITIES MATRIX: Construct a Markdown table extracting all critical figures, statistics, key dates, monetary values, or key stakeholders referenced across the themes.
5. ACTIONABLE IMPLICATIONS / DECISION IMPACT: Provide 3-4 concrete takeaways or next steps directly tailored to my stated goal above, answering: "Based on these themes, what should I do next?"
TONE & EXECUTION GUIDELINES:
* Tone: Highly analytical, objective, precise, and rigorous.
* Accuracy Safeguard: Rely strictly on the information provided in the text. Do not invent or extrapolate unverified facts.
* Readability: Use bold headers, bulleted lists, and clean Markdown tables to make the analysis instantly scannable for executive review.
Here is the document to analyze:
[PASTE YOUR DOCUMENT TEXT HERE]Is this your AI Workout of the Week (WoW)? Cast your vote!
That's all we've got for you today.
Did you like today's content? We'd love to hear from you! Please share your thoughts on our content below👇
What'd you think of today's email?
Your feedback means a lot to us and helps improve the quality of our newsletter.
