Hollywood heavyweights sue Midjourney
PLUS: Patients control AI and robotics with thought
Zach Mink
June 12, 2025
Read Online | Sign Up | Advertise
Good morning, AI enthusiasts. The Mouse just lawyered up — and unlike media outlets that settled for licensing deals, Disney, Marvel, and other major Hollywood studios are choosing the courtroom over compromise.
With Hollywood playing hardball on Midjourney’s reproduction of its IP in outputs and model training, the AI industry’s fair use defense may be about to get its toughest stress test yet.
In today’s AI rundown:
Disney, Universal sue Midjourney over copyright
TBC goes all-in on AI with Dia browser
How to connect Claude to external applications
Meta's new AI learns real-world physics
4 new AI tools & 4 job opportunities
LATEST DEVELOPMENTS
MIDJOURNEY
👨🏻⚖️ Disney, Universal sue Midjourney over copyright
Image source: Legal filing
The Rundown: Hollywood giants, including Disney and Universal, just filed a lawsuit against AI image generation startup Midjourney, alleging that the company’s models are a massive engine for copyright theft.
The details:
The lawsuit marks Hollywood's first major legal action against AI companies, with Disney, Marvel, Lucasfilm, Universal, and others joining forces.
The group claims that Midjourney built its models by scraping the studios' iconic characters, allowing users to generate endless unauthorized knockoffs.
Dozens of examples were included in the complaint, featuring side-by-side comparisons of characters like Yoda, Shrek, Spider-Man, and Minions.
Disney’s legal counsel said that while the company is ‘bullish’ on the promise of AI, “piracy is piracy.”
Why it matters: While many major media outlets have struck lucrative AI licensing deals for their content, Hollywood studios are playing hardball. This case could have major ramifications for AI companies’ “fair use” claims — potentially setting legal precedents that could impact not just Midjourney but the entire industry.
TOGETHER WITH VANTA
💨 The compliance platform that keeps up with AI speed
The Rundown: While you're sprinting toward your next round or first enterprise contract, buyers are demanding security, compliance, and transparency from day one. Vanta bridges that gap with compliance automation built specifically for AI's unique challenges.
How Vanta accelerates AI startups:
Automated audit readiness that evolves with your models, infrastructure, and customers
Continuous security monitoring designed for AI’s rapid iteration cycles
Used by AI innovators like LangChain and Factory, who turned early compliance into deal-closing credibility
Get compliant at AI-speed with Vanta — book a demo and claim $1,000 off.
THE BROWSER COMPANY
🌐 TBC goes all-in on AI with Dia browser
Image source: The Browser Company
The Rundown: The Browser Company released its AI-first Dia browser in beta, featuring an integrated chatbot that can see every tab, take agentic actions, and adapt to user preferences and patterns.
The details:
Dia integrates its AI directly into the URL bar, allowing users to chat with their open tabs, get summaries, and draft content without leaving their workflow.
Dia’s chatbot can analyze multiple tabs at once, draft emails based on a user’s writing style, and use days of browsing history for personalized responses.
It uses a system of “Skills” or specialized AI agents tailored for specific tasks like shopping or coding that remember context from relevant tabs.
Beta access launched today for existing Arc users on Mac, with all data encrypted locally and wiped from servers immediately after processing.
Why it matters: The web is trending towards AI and agentic capabilities — and the race to create the best “AI-native” browser is officially on. Dia feels directly correct in that AI will evolve to where users are instead of being isolated in platforms, but the likes of Google and OpenAI are undoubtedly coming with powerful solutions of their own.
AI TRAINING
🔗 How to connect Claude to external applications
The Rundown: In this tutorial, you will learn how to use Claude's new MCP integration to connect AI directly to thousands of external tools like Google Docs, Slack, WhatsApp, and more via Zapier's MCP server.
Step-by-step:
Go to Claude Settings and look for “Add integrations” in the “Search and tools” option
Visit the Zapier MCP site, create a free account, and add Claude in “New MCP Server”
In the Configure tab, click “Add tool” and search for apps like Google Docs, or Slack
Copy the Integration URL from the Connect tab, return to Claude, and paste it in “Add integrations”
Test it! Ask Claude: "Create a new Google Doc about [topic]" and watch it work across apps
Pro tip: Start with 1-2 essential tools, test thoroughly, then add more apps as needed. You can edit or disable tools anytime.
PRESENTED BY SPEECHMATICS
🎙️ The voice AI that protects conversations that matter
The Rundown: Most Voice AI Agents send speech data to the cloud — yours doesn't have to. Speechmatics lets you deploy directly on-device, on-prem, or in-cloud, safeguarding patient confidentiality and customer trust without sacrificing performance.
Speechmatics’ enterprise-grade capabilities include:
Deployment flexibility across on-device, on-premises, hybrid, or cloud environments
Proven accuracy across 55+ languages with multilingual conversation handling
Infrastructure that handles 500+ years of audio monthly for governments and Fortune 500 companies
META
🎢 Meta's new AI learns real-world physics
Image source: Midjourney
The Rundown: Meta released V-JEPA 2, a “world model” that gives AI systems the ability to understand physics and predict real-world outcomes — enabling robots to navigate unfamiliar environments and manipulate objects they've never seen before.
The details:
The 1.2B parameter model was trained on 1M+ hours of video, learning how objects move, interact, and respond to actions in the physical world.
V-JEPA 2 achieved 65-80% success rates in picking and placing unfamiliar objects in new environments, using visual goals to plan multi-step tasks.
Meta claims the model runs 30x faster than Nvidia's competing Cosmos model while achieving SOTA performance on video understanding benchmarks.
The company also released three new benchmarks revealing that while humans score 85-95% on physical reasoning tasks, current AI models struggle.
Why it matters: Grounding AI in physical reality instead of text-based reasoning is an important development as both AI agents and robots become more and more utilized in real-world tasks. Meta's approach can help AI adapt to the messy, unpredictable real-world environments on the fly without requiring massive task-specific training datasets.
QUICK HITS
💼 AI Job Opportunities
🎨 The Rundown - Designer (Brand & Platform)
⚡ xAI - Electric Engineer
🛡️ Horizon3 - Attack Engineer
✍️ Abridge - Technical Writer
📰 Everything else in AI today
OpenAI CEO Sam Altman revealed that the company’s first open-weight model, expected in June, will take “a little more time” but be “very, very worth the wait.”
Apple execs defended their AI efforts in an interview with the WSJ, saying the company made the right call to not ship AI Siri that didn’t meet quality standards.
Meta announced new AI video editing capabilities in its Meta AI app, allowing users to quickly change outfits, locations, lighting, and more with preset prompts.
Mistral launched Mistral Compute, an AI stack offering GPU access, orchestration, and model training services, positioning itself as an alternative to cloud giants.
Windsurf unveiled the Windsurf Browser, giving its Cascade agentic coding assistant full awareness of web activity for in-context support.
Starbucks is piloting Green Dot Assist, a new AI tool to help baristas answer questions and access guidance in real-time.
Midjourney launched video ranking, allowing users to explore and rate outputs from its soon-to-be-released video model.
COMMUNITY
🎥 Join our next live workshop
Join our next workshop this Friday, June 13th, at 4 PM EST with Dr. Alvaro Cintas, The Rundown’s AI professor. By the end of the workshop, you’ll know how to get significantly more value out of whichever model you use, whether it’s o3 or o3-Pro.
RSVP here. Not a member? Join The Rundown University on a 14-day free trial.
See you soon,
Rowan, Joey, Zach, Alvaro, and Jason—The Rundown’s editorial team
Stay Ahead on AI.
Join 1,000,000+ readers getting bite-size AI news updates straight to their inbox every morning with The Rundown AI newsletter. It's 100% free.