The Rundown AI / Articles / AI / Disney's billion-dollar AI bet on OpenAI

Disney's billion-dollar AI bet on OpenAI

PLUS: Patients control AI and robotics with thought

Zach Mink

December 12, 2025

Good morning, AI enthusiasts. Disney just did what every media company has been reluctant to do — hand over the keys to its kingdom to AI.

With a new licensing deal and $1B investment into OpenAI, the Mouse is making its first (legal) appearance in Sora — and hitting the AI leader’s rivals with cease-and-desists in the process to completely shift the IP landscape.

Reminder: Our next live workshop is today at 4 PM EST. Attend and learn how to leverage Google’s Workspace Studio to create practical AI automations for tasks like email inbox triage, template building, and more. RSVP here.

In today’s AI rundown:

OpenAI, Disney lock in $1B licensing deal
OpenAI drops GPT-5.2, hits back at Google
Create and deploy voice agents for your business
Google opens upgraded Deep Research agent to devs
4 new AI tools, community workflows, and more

LATEST DEVELOPMENTS

OPENAI & DISNEY

🤝 OpenAI, Disney lock in $1B licensing deal

Image source: Reve / The Rundown

The Rundown: Disney just announced a three-year licensing deal with OpenAI, giving Sora users access to over 200 characters from Disney, Marvel, Pixar, and Star Wars — alongside a $1B equity investment into the AI leader.

The details:

Fans will be able to use Disney-owned IP like Mickey Mouse, Darth Vader, and the Avengers in video generations, with select creations streaming on Disney+.
Disney is also deploying OpenAI’s APIs across products and rolling out ChatGPT internally as part of a broader enterprise push.
The deal specifically excludes talent likenesses and voices, sidestepping the more complex IP battles still playing out across Hollywood.
Disney also sent a cease-and-desist to Google on the same day, accusing the OAI rival of generating unauthorized Disney content at “massive scale”.

Why it matters: CEO Bob Iger made waves last month when he said AI was coming to Disney+, but this is a bigger move than anyone imagined. For OAI, the deal gives them a unique leg up on utilizing popular IPs without legal trouble — and leads to an even more aggressive enforcement against rival generators as they protect that advantage.

TOGETHER WITH ATLASSIAN ROVO

👋 Meet Rovo, AI that knows your business

The Rundown: Discover Atlassian Rovo - AI that knows your business. Rovo connects teams, knowledge, and tools so you can move faster and work smarter – together.

Why Rovo?

Rovo connects to all your favorite SaaS apps
Rovo brings organizational knowledge and context into every workflow
It’s already built into Jira, Confluence, and more.
And it's built on Atlassian’s enterprise-grade security & privacy.

Try Rovo today.

OPENAI

🧄 OpenAI drops GPT-5.2, hits back at Google

Image source: OpenAI

The Rundown: OpenAI released its GPT-5.2 model family, calling it the company’s “most capable series yet for professional knowledge work”, arriving just weeks after an internal memo warned the company was losing ground to Google’s Gemini lineup.

The details:

The release comes in three tiers: Instant for quick queries, Thinking for complex reasoning tasks, and Pro for maximum accuracy on hard problems.
5.2 is an upgrade across benchmarks from 5.1, showing notable gains in hallucination rate, vision, coding, long-context reasoning, and tool use.
On GDPval, GPT-5.2 Thinking beat or matched industry professionals 71% of the time across real-world tasks like spreadsheets and presentations.
5.2 was reportedly released despite internal requests to delay for more polish, with OAI pushing ahead as Gemini 3 topped most public leaderboards.

Why it matters: OpenAI’s codenamed ‘Garlic’ model is here, and it packs an impressive counter-punch to Gemini 3 despite feeling like a rushed production. This month felt like one of the first times OAI has been on its heels — but in addition to 5.2, it likely has some buzzy Christmas drops up its sleeve in the coming weeks.

AI TRAINING

🗣️ Create and deploy voice agents for your business

The Rundown: In this tutorial, you will learn how to use Cartesia to build and deploy AI voice agents that can handle calls, take orders, or answer customer questions using natural, human-like speech.

Step-by-step:

Go to Cartesia, click "Start for Free", explore the dashboard, and test voices in "Text to Speech" using Sonic 3.0 for best latency and clarity
Scroll to "Voice Agents", click "Text to Agent", describe your agent (e.g., "Pizza order assistant that greets customers, takes orders, confirms details, calculates totals")
Select voice preference, click "Generate", then test using the dialer at +1 (515) 800-8360 - verify response speed, order handling, and voice clarity
Click "Promote to Production" to get a working number, publish it on the site, and monitor performance in "Metrics" (calls handled, duration, credit usage)

Pro tip: Refine prompts as needed based on customer interactions.

PRESENTED BY MATILLION

🔧 95% of AI projects fail — here’s the fix

The Rundown: Fact – most AI projects stall before delivering value. Leading CDOs are turning to Maia, the agentic data team, to cut manual work by 90%, launch initiatives 10x faster, and finally clear backlog bottlenecks.

Download this ebook to learn how to:

Cut delivery times
Scale data work without extra hires
Take AI projects from PoC to production

Get the ebook.

GOOGLE

🧠 Google opens upgraded Deep Research agent to devs

Image source: Google

The Rundown: Google just released a significantly upgraded version of its Deep Research agent, now available to developers through a new Interactions API — with consumer rollouts coming soon to Search, NotebookLM, and the Gemini app.

The details:

The agent runs on the new Gemini 3 Pro, iteratively planning searches, reading results, identifying gaps, and querying again until reaching an answer.
A new Interactions API unifies access to Gemini models and agents, with native support for MCP connections to external tools and data sources.
Google also open-sourced DeepSearchQA, a 900-task benchmark designed to test multi-step web research more rigorously than existing evaluations.
Google claims SOTA scores on Humanity's Last Exam (46.4%) and the DeepSearchQA benchmark (66.1%), outpacing the Gemini 3 Pro base model.

Why it matters: There’s no shortage of competition in the Deep Research category, but Google’s updated agent builds nicely on top of the already powerful Gemini 3 releases. It also provides devs with the opportunity to leverage the tool in their own apps for the first time, bringing an upgrade to the research layer of third-party builds.

QUICK HITS

🛠️ Trending AI Tools

💻 Augment Code Review - Purpose-built code review that catches critical issues without the noise*
📚 Gemini Deep Research Agent - SOTA agent for long synthesis tasks
🤖 GPT-5.2 - OpenAI’s new most advanced frontier model family
🏷️ Shopify SimGym - Simulate buyer behavior with AI shoppers

*Sponsored Listing

📰 Everything else in AI today

TIME Magazine named “the architects of AI” as its 2025 Person of the Year, spotlighting leaders Jensen Huang, Sam Altman, Dario Amodei, and Elon Musk.

Cursor unveiled a new visual editor that lets developers drag, drop, and rearrange interface elements while AI agents automatically update the underlying code.

Shopify dropped new Winter '26 AI features, including SimGym for simulating shopper behavior and Agentic Storefronts that surface products in AI platforms.

Runway introduced GWM-1, its first "General World Model" that can simulate interactive, explorable environments in real time.

Google Labs launched Disco, an experimental browser that uses Gemini 3 to generate custom web applications based on your open tabs and browsing tasks.

ElevenLabs announced a new partnership with Meta, bringing its audio and voice tech to creators on Instagram, Horizon, and more.

COMMUNITY

🤝 Community AI workflows

Every newsletter, we showcase how a reader is using AI to work smarter, save time, or make life easier.

Today’s workflow comes from reader Jyothi V. in India:

"I used AI as an unexpected consumer-rights strategist. When a store refused returns on a 13-item order, I simply asked how to contact them. The AI analyzed the situation, identified that the policy was buried, and suggested a negotiation strategy referencing consumer rights. It drafted a firm escalation message, and the store ended up accepting my returns. It felt like having a built-in advocate who sees angles I miss."

How do you use AI? Tell us here.