This week, Disney just made its biggest AI statement yet by partnering with OpenAI and putting iconic characters inside Sora under a licensed framework. Days later, President Trump signed an executive order aimed at reducing the growing patchwork of state AI laws. Let’s dive in ⬇️

In today’s newsletter ↓
🍿 A studio sized bet on licensed AI video
🏛️ A federal push to overrule state AI laws
⚡ GPT 5.2 enters the model race
🧩 Claude Code lands inside Slack
🧭 Weekly Challenge: Build a reusable AI brief

Disney and OpenAI just announced a three year licensing agreement that brings more than 200 characters from Disney, Marvel, Pixar, and Star Wars into Sora for short fan prompted videos, plus the same licensed set for ChatGPT Images.

This is the part that matters. Disney is not just letting creators play. It is normalizing licensed generative video in the most mainstream IP catalog on the planet. If this works, other rights holders will be under pressure to offer licensed options instead of relying on lawsuits and takedowns.

The agreement frames the output as short social videos created from user prompts, with a selection of fan made clips eligible to stream on Disney Plus. Disney also describes strict boundaries that focus on animated, masked, and creature characters, alongside costumes, props, vehicles, and environments, while excluding real actor likenesses and voices. Those details appear in the OpenAI and Disney announcements of the partnership.

That carve out is not a footnote. It is the entire legal and labor bridge. It protects performers, reduces rights conflicts, and gives Disney a clearer line to defend if the internet tries to push the tool into impersonation.

The market read this as a de risked path to AI experimentation inside a premium content machine. Another angle in analysis of the IP and platform implications is that this structure could become a blueprint for how big IP holders get paid without endorsing unlicensed training.

Some investors see faster iteration in early production work like storyboards, pre-visualization, and internal prototypes. Others worry that AI will encourage volume over craft and dilute brand trust. That split shows up in reporting on the investor debate.

The more subtle point is leverage. If Disney can funnel creators into a licensed pipeline, it can turn generative video from a piracy headache into a new distribution and engagement layer.

For years, studios have tried to stop unlicensed training and lookalike outputs. This deal flips the posture. Instead of forcing courts to draw the line, Disney is drawing a commercial line and inviting fans inside it. Commentary on how studios may move from lawsuits to licensing highlights why this could become the template for other major catalogs.

The real test will be whether safety controls actually hold up at scale and whether Disney can curate without turning the product into a bland creativity sandbox.

President Trump signed an executive order titled Ensuring a National Policy Framework for Artificial Intelligence. The order aims to discourage state by state AI regulation and push the federal government to take the lead on AI policy and enforcement. You can read the full executive order text.

This does not instantly erase state laws. But it signals a willingness to fight them and it creates a policy direction that agencies and courts could use as cover.

The order calls for a coordinated approach led by the federal government and frames state AI laws as potential obstacles to national AI leadership. The White House fact sheet on the national AI framework lays out the administration’s argument and the broad intent.

Reporting also notes that the order directs agencies to look for ways to challenge or discourage state measures and to build an enforcement posture that supports a minimally burdensome national framework. A detailed breakdown of the directive and its likely limits explains why this may create uncertainty before it creates clarity in reporting on the one rulebook order.

States have been moving faster than Congress on deepfakes, hiring tools, consumer protections, and child safety. Coverage of the order’s attempt to block state laws highlights how uncertain the preemption question still is in practice. A single federal approach could reduce compliance chaos for companies, but it could also slow protections down if federal standards arrive late or stay vague.

🏢 Microsoft doubles down on global AI infrastructure

🔌 Washington adds bias testing to federal AI contracts

🧠 OpenAI escalates the model race with GPT 5.2

🧩 Anthropic brings code writing directly into Slack

⚙️ Adobe points to AI as a real revenue driver

🌡️ Researchers warn reasoning models carry steep energy costs

Most people use AI like a slot machine. This challenge turns it into a repeatable work tool.

Here’s what to do:

🧩 Step 1: Define the structure Create a single prompt that forces structure. Ask for five bullets only. Each bullet must include one key fact, one risk, and one next step.

🔁 Step 2: Run it consistently Run the same prompt on three real inputs this week. Do not change the wording between runs.

📊 Step 3: Score the output Score each output from 0 to 2 on three criteria.

🎯 Accuracy Did it match the source text without inventing details

🛠️ Usefulness Would you act on it without more digging

🧠 Clarity Could you forward it to someone with zero context

✍️ Step 4: Refine once Rewrite the prompt one time based on what failed most. Save the final version as your default brief.

If your brief does not save at least ten minutes per run, it is not a workflow yet.

From Disney and OpenAI partnering to the U.S. federal government forcing states to comply with a nationalized AI agenda, the world of AI never sleeps. Hit reply and let us know your thoughts.

Zoe from Overclocked