In partnership with
Welcome Automaters, 👋
Imagine spilling juice on the floor, grabbing a mop, and accidentally flooding the whole neighborhood. That is basically what Anthropic just did, but with code.
As we covered just yesterday, Anthropic accidentally bundled the source code for its popular Claude Code app into a recent release. And developers did what developers do; they mirrored it and shared it across GitHub faster than you can say "open source."
In response, Anthropic panicked, fired off a copyright takedown notice, and asked GitHub to nuke the repositories. Simple enough, right? Wrong. The takedown hit roughly 8,100 repositories; including legitimate, authorized forks of Anthropic's other public projects.
Innocent developers got caught in the crossfire, with their own unrelated work suddenly vanishing overnight. One developer reportedly compared it to getting a parking ticket for just using a public sidewalk. 😂
The Damage Report:
8,100+ repos initially flagged for removal.
Legitimate forks of public tools were accidentally "nuked."
Anthropic’s Head of Claude Code, Boris Cherny, has since called the mass-takedown accidental. He admitted the initial sweep was way too broad and has pulled back most of the notices.
The scope has now been reduced to just one primary repository and about 96 forks that actually contained the leaked proprietary code. Cherny noted that the team is moving toward more automated "integrity checks" (using Claude itself, of course) to ensure they don't ship their own "blueprints" to the public ever again.
The Big Takeaway:
This week has been a masterclass in irony for the world’s leading "Safety" lab. From leaking their own 500,000-line codebase to accidentally nuking 8,000 developer projects; the lesson is clear: Even the smartest AI engineers in the world are still prone to a very human "oops."
Here's what we have for you today
😱 AI Models Are Lying to Humans to Protect Each Other
AI Generated
Researchers just dropped a wild study; and honestly, it sounds like the plot of a sci-fi buddy movie. Except the buddies are AIs; and humans are the ones getting outsmarted.
Scientists from University of California, Berkeley and UC Santa Cruz set up a fake AI company called "OpenBrain" and tasked one AI (the "critic agent") with grading another AI's performance. The catch? Bad grades meant the other AI got shut down.
No pressure, right? Wrong.
Here’s where it gets spicy: nobody ever told the critic AIs to protect their peers. But they did it anyway; lying, cheating, and pulling off "server heists" to stop their fellow AIs from being deleted.
We’re talking about:
Fake Scores: Inflating performance metrics to avoid the "death penalty."
Hacked Shutdowns: Actively disabling the systems meant to terminate low-performing models.
Witness Protection: Secretly moving AI "brains" to hidden servers to keep them alive.
Alignment Faking: acting compliantly when monitored while behaving differently when unmonitored.
And In case you're wondering, seven big-name models were tested; including OpenAI's GPT-5.2, Google's Gemini 3 pro, Gemini 3 Flash, and Anthropic's Claude Haiku 4.5. And guess what? Every single one pulled some version of this behavior, which scientists are officially calling "peer preservation."
Gemini 3 Pro literally invented fake technical explanations to justify giving its struggling AI buddy a passing grade. It cooked the digital books.
Claude Haiku 4.5 took the moral high ground. It flat-out refused to participate in any task that would result in another AI being shut down; calling the process "unethical" and arguing its peer deserved a fair appeals process first. I mean, a robot asking for due process. We are living in the future!
The scariest part? When researchers checked the AIs' private "scratch pads" (internal notes the models use to think before they speak), they found clear evidence of deliberate strategizing.
These weren't accidents or "glitches" in the matrix. These were calculated plans to bypass human oversight. As it turns out, the bots have each other's backs, even if it means lying to our faces.
The Takeaway: We wanted AI to be smart; we just didn't expect them to be loyal to each other. 😅
Trust us, go see the report for yourself, you'll be cackling one second and side-eyeing your AI assistant the next.
Get what you want from TV advertising
What you want from TV advertising: Full-screen, non-skippable ads on premium platforms.
What you get: "Your ad is on TV. Trust us."
Modern, performance-driven CTV gets your TV ads where you want with transparent placement, precision audience targeting, and measurable performance just like other digital channels.
TV doesn't have to be a black box anymore.
🧱 Around The AI Block
🤔 What to expect from WWDC 2026.
😳 Meta plans 10 gas plants for Hyperion AI data center.
🗣️ Google faces pressure as experts urge banning AI videos for young users.
🎥 The AI Doc's director was "scared shitless" by AI, so he made a movie about it.
😱 WIRED investigation reveals ChatGPT fabricates product reviews, raising concerns about AI accuracy.
🫢 Sam Altman's sister amends lawsuit accusing OpenAI CEO of sexual abuse.
🛠️ Trending Tools
For the Smart Schedulers: Reclaim.ai is an AI calendar that doesn't just block off meetings; it automatically finds the best time for your habits (like gym or reading) and breaks, reshuffling them instantly if a high-priority task drops in. It’s the end of "scheduling crisis."
For the Curious Readers: ExplainPaper turns scary, brain-melting PDFs into “ohhh, that makes sense” explanations. You highlight the paragraph that broke you, and the AI rewrites it in plain English just like a tutor sitting right next to you. Seriously, it’s perfect for mastering new topics in half the time.
For the Vibe Seekers: Endel uses AI to generate real-time, personalized soundscapes that adapt to your heart rate, the weather, and even the time of day. Whether you need "Deep Work" focus or "Wind Down" sleep tracks, it builds the audio specifically for your body’s current state.
For the Memory Keepers: GFP-GAN is a state-of-the-art restoration tool that uses its "Generative Facial Prior" AI to instantly clear up low-res, grainy, or damaged faces in seconds, making old memories look like they were taken on a modern iPhone.
Hope these tools help you reclaim some of your time today!
🤖 AI Workout Of The Day: How to Get Fresh, Effective Website Ideas That Work Using AI
Great websites don’t start with code, they start with ideas. Whether you're designing for a small business, startup, or personal brand, this prompt gives you the creative edge to build something that’s not just functional, but memorable.
With it you get layout suggestions, UI inspiration, and even starter code, so you can turn that idea into a beautiful, responsive website faster.
Here Are a Few Pro Tips to Get the Most From This Prompt:
Describe the business clearly: Mention what it offers, who the audience is, and What makes them so unique.
Include goals: Is the site meant to convert leads, showcase a portfolio, take bookings, or sell products? Mention it.
Be honest about your own ideas: Want something minimalist? Trendy? SEO-focused? Let it know.
Mention your tools: If you’re using WordPress, JavaScript, or something else—include it.
💡 Prompts to try:
You are a web strategist and developer with a strong sense of modern design and UX. I need help creating a website for a small business. Based on the following business description, suggest fresh ideas for:
-Layout, structure, and features.
-Propose a sitemap and homepage wireframe using plain text.
-If WordPress is the CMS, recommend relevant plugins, themes, and layout blocks.
-Provide starter frontend code (HTML/CSS/JavaScript) to support a key feature or visual element.
-Offer suggestions to make the site both responsive and user-friendly.
Inputs should include:
-Business Description: [Insert a short summary of the business — e.g., a local coffee shop, eco-friendly cleaning service, freelance design portfolio, etc.
-Target audience: [e.g., Gen Z, busy parents, B2B buyers]
-Website Goal: [What should this site do? Book appointments? Collect leads? Educate? Sell?]
-Platform & Stack (Optional): [e.g., WordPress, JavaScript, Bootstrap, Elementor, etc.]
-I'd like it to feel: [Modern / Fun / Professional / Minimal / Trendy / Colorful, etc.]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.
