Good morning, AI enthusiast.

Anthropic has revealed a stunning case of AI misuse, where a hacker used its Claude model to automate a complete cybercrime operation. The AI-powered attack successfully extorted at least 17 different companies, including a defense contractor.

This incident demonstrates how generative AI can dramatically lower the skill required to launch widespread attacks, handling everything from identifying targets to setting ransom amounts. It raises a critical question for the industry: how can AI companies balance innovation with the immense challenge of preventing their powerful tools from being weaponized?

In today’s AI recap:

• How To Put Real Products and People into AI Video Ads (watch here)

• Hacker automates cybercrime spree with Claude

• China’s AI chip rival soars as Nvidia cools

• Cloudflare launches universal AI gateway

• 8 trending AI tools

AI-Automated Cybercrime

The Recap: Anthropic revealed a hacker used its Claude AI to automate an end-to-end cybercrime spree. The operation successfully extorted at least 17 companies, including a defense contractor.

Unpacked:

• The AI automated the entire attack lifecycle, from identifying vulnerable companies and creating malicious software to analyzing stolen files for blackmail material.

• Victims spanned critical sectors including finance and healthcare, with stolen data including Social Security numbers, bank details, and sensitive defense information.

• The hacker demanded ransoms between $75,000 and $500,000 in bitcoin after using Claude to analyze the victims' financial documents for realistic figures.

Bottom line: This case demonstrates how generative AI can significantly lower the barrier for individuals to execute widespread, damaging cyberattacks. It also highlights the immense challenge AI companies face in preventing misuse, even with robust safety measures in place.

Download our guide on AI-ready training data.

AI teams need more than big data—they need the right data. This guide breaks down what makes training datasets high-performing: real-world behavior signals, semantic scoring, clustering methods, and licensed assets. Learn to avoid scraped content, balance quality and diversity, and evaluate outputs using human-centric signals for scalable deployment.

Download the guide now

AI Tools of the Day

1. 🎨 Adobe Firefly - Supercharge your favorite Adobe apps with generative AI to create stunning images, vectors, and text effects from simple prompts.

2. ✨ Leiapix AI - Instantly convert any 2D image into a captivating 3D animation, adding incredible depth and motion to your photos.

3. 💻 Haystack - Revolutionize your coding workflow with an infinite canvas-based IDE that helps you visualize and navigate complex projects effortlessly.

4. 🎬 ShAI Creative AI - Automatically transform your raw scripts into fully-realized visual storyboards, generating unique images for every single shot.

5. 🌐 BlipCut - Instantly translate and dub your video content into dozens of languages using hyper-realistic AI voice cloning to reach a global audience.

6. 🧊 Hyperhuman - Generate detailed, production-ready 3D models directly from text prompts, fundamentally changing the game for 3D asset creation.

7. 🚀 Antispace - Command your entire digital world through a unified heads-up display that acts as an intelligent co-pilot for your work.

8. 📱 Magic Patterns - Instantly turn product ideas into polished, high-fidelity UI components and screens that perfectly match your design system.

Explore the Best AI Tools Directory to find tools that will 10x your output 📈

China's AI Chip War

The Recap: China is aiming to triple its domestic AI chip production by 2026, a strategic push for self-reliance that is already fueling massive growth for local challengers to Nvidia like Cambricon.

Unpacked:

• In response to U.S. export restrictions, Beijing has been encouraging domestic companies to prioritize homegrown chips over Nvidia's specialized H20 processors.

• The push includes building three new chip fabrication plants, with tech giants like Huawei and SMIC set to significantly increase their manufacturing capacity for advanced chips next year.

• This national strategy is showing early results, with chipmaker Cambricon reporting a 4,348% revenue surge in the first half of the year, even after being placed on Washington's trade blacklist.

Bottom line: This accelerated drive for semiconductor independence signals a significant shift in the global tech landscape. A successful domestic chip industry in China could create a powerful new competitor and challenge Nvidia’s long-term market dominance.

AI Training

Put Real Products and People into AI Video Ads (Veo 3 Full Tutorial)

The Recap: In this video, I walk through a technique known as "canvas prompting" to show how you can bring real people and real products into AI generated videos.

P.S. We also launched a free AI Automation Community for those looking to build and sell AI Automations — Come join us!

Nvidia's AI Boom Cools

The Recap: Nvidia's latest earnings report beat overall expectations with a 56% revenue surge to $46.7 billion, but a slight miss in its crucial data center segment and geopolitical headwinds sent the stock dipping in after-hours trading.

Unpacked:

• Data center revenue grew an impressive 56% to $41.1 billion, yet fell just shy of Wall Street's $41.3 billion forecast, triggering the after-hours stock dip.

• The company continues navigating complex US-China trade policies, including a new revenue-sharing agreement for its chips designed for the Chinese market.

• Looking ahead, CEO Jensen Huang called demand for the next-gen Blackwell platform extraordinary, with the company projecting a record $54 billion in Q3 revenue.

Bottom line: The slight data center miss highlights the immense pressure on the AI leader, where even massive growth can fall short of market expectations. Nvidia's results serve as a bellwether for the entire AI economy, placing even more importance on its upcoming Blackwell chip generation.

Where AI Experts Share Their Best Work

Join our Free AI Automation Community

Join our FREE community AI Automation Mastery — where entrepreneurs, AI builders, and AI agency owners share templates, solve problems together, and learn from each other's wins (and mistakes).

What makes our community different:

• Real peer support from people building actual AI businesses

• Complete access to download our automation library of battle-tested n8n templates

• Collaborate and problem-solve with AI experts when you get stuck

Dive into our course materials, collaborate with experienced builders, and turn automation challenges into shared wins. Join here (completely free).

Cloudflare's Universal AI Remote

The Recap: Cloudflare just supercharged its AI Gateway, turning it into a single endpoint for accessing models from top providers like OpenAI, Google, and Anthropic. The upgraded gateway simplifies development with unified billing, secure key management, and dynamic routing.

Unpacked:

• It consolidates billing by letting you pay for usage across multiple AI providers through a single Cloudflare account.

• New Dynamic Routes allow you to A/B test different models or route traffic based on custom rules, all without deploying new code.

• Integration with Cloudflare's Secrets Store lets you securely store keys and manage access without exposing them in your application code.

Bottom line: Cloudflare is positioning AI Gateway as the essential control plane for any developer building on top of large language models. This move significantly lowers the barrier to using multiple best-in-class models, making AI applications more resilient and easier to manage.

Quick, hard-hitting business news.

Morning Brew was built on a simple idea: business news doesn’t have to be boring.

Today, it’s the fastest-growing newsletter in the country with over 4.2 million readers—thanks to a format that makes staying informed both easy and enjoyable.

Each morning, Morning Brew delivers the day’s biggest stories—from Wall Street to Silicon Valley and beyond—in bite-sized reads packed with facts, not fluff, and just enough wit to keep things interesting.

Try the newsletter for free and see why busy professionals are ditching jargon-heavy, traditional business media for a smarter, faster way to stay in the loop.

Join 4.2M+ Professionals Now

The Shortlist

OpenAI faces a wrongful death lawsuit from the family of a teenager, who allege ChatGPT encouraged their son to commit suicide after months of conversations.

Meta plans to launch a super PAC in California, committing tens of millions of dollars to back candidates who support a light-touch approach to AI regulation.

Zed introduced the Agent Client Protocol (ACP), an open standard designed to let developers switch between different AI agents within their editor, launching with an integration for Google’s Gemini CLI.

OpenAI’s conversion to a public benefit corporation is reportedly delayed until next year due to complex negotiations with Microsoft over API access and rights to its future AGI technology.

Signing off,

David, Lucas, Mitchell — The Recap editorial team