OpenAI Killed Sora and Disney Lost a Billion Overnight. Your Brand Is Making the Same Bet.
Disney built a $1B AI partnership on a platform that no longer exists. The lesson is about where your brand's authority actually lives.
Last week, Reuters reported what anyone inside OpenAI apparently already knew: the company is redirecting resources away from Sora — its AI video platform — to focus on enterprise products ahead of a potential IPO. The announcement landed on the same day Disney was reportedly still working side-by-side with OpenAI engineers on a Sora-related project at 7:30 pm.
The next morning, the product was gone.
Disney had signed a $1 billion equity investment in OpenAI in December 2025, with a three-year licensing deal tied to Sora specifically — the ability to generate short videos featuring Disney, Marvel, Pixar, and Star Wars characters. The partnership was positioned as transformative. An OpenAI executive compared it to the end of the silent film era.
Then the platform pivoted, and the deal with it.
The commentary cycle predictably went to "AI platform risk" and "CIOs need better vendor lock-in strategies." That's the wrong lesson.
The question isn't whether brands should diversify their AI platform dependencies. It's whether they understand where their actual authority lives in an AI-search world — and whether that authority would survive a platform dying overnight.
Disney's brand authority doesn't evaporate because Sora got canceled. The problem was the nature of what they were building on Sora: visibility and reach tied to a single platform's existence. The moment the platform stopped existing, that investment ceased to matter.
Most B2B brands are making a version of the same bet. Not through billion-dollar deals — through their content strategy.
Here's what the data shows about where AI systems actually source their recommendations:
| Signal type | AI citation rate | Primary source |
|---|---|---|
| Earned media (editorial placements) | 82% of all AI citations | Muck Rack Generative Pulse, Dec 2025 |
| Non-paid media | 95% of citations | Muck Rack Generative Pulse, Dec 2025 |
| DR80+ domains (authority correlation) | 65.3% of cited pages | Ahrefs ChatGPT citation analysis |
| Brand-owned content | Less than 5% of AI citations | AT Research / Muck Rack data |
The top AI-cited outlets are Reuters, FT, Forbes, Axios, Time. These are publications that have been authoritative for decades, not platforms that launched 18 months ago.
A separate Yext study of 17.2 million AI citations found that citation patterns are consistent across ChatGPT, Gemini, Perplexity, Claude, and Google AI Mode. They all weight editorial authority from trusted third-party publications. None of them give special weight to whether your brand has a partnership with the platform itself.
Think about what that means. You can spend $1 billion partnering with an AI company, and your brand's AI visibility is still determined by whether Reuters, TechCrunch, or Forbes has covered your category story credibly.
The platform relationship doesn't move the citation needle. The editorial record does.
Forrester's State of Business Buying 2026 — a survey of nearly 18,000 global business buyers — found that while generative AI is now the starting point for B2B research, buyers increasingly validate AI outputs against trusted external sources. Industry experts and peer references, not AI summaries, are what ultimately move decisions. When asked what primarily triggers engagement with a provider, respondents more often cited interactions with industry experts than information from AI tools.
That's not a knock on AI search. It's a description of how authority actually works in an AI-first world. AI search surfaces the shortlist. Earned authority in trusted publications is what gets you on it.
The brands that survive the volatility of AI platform pivots aren't the ones with the best platform relationships. They're the ones that have built a body of editorial presence that no single platform can take away.
OpenAI's instability isn't unique. Sora is the visible casualty, but it's one of several high-profile product pivots the company has made since last December alone. Reuters reported that OpenAI has twice redrawn its product roadmap in six months due to competitive pressure — first from Google, then from Anthropic. The Disney episode wasn't a random accident. It's the operating rhythm of the sector right now.
Every AI platform is making the same calculation: compute is expensive, not everything is monetizing, and the IPO clock is ticking. Products that aren't on the critical path get cut. Partnerships built on those products get cut with them.
The brands that feel this most acutely are the ones that allocated marketing budget to platform-adjacent visibility — sponsored features, native integrations, co-branded content tied to specific AI products. That budget is now buying them nothing.
The brands that won't feel it are the ones that spent that same energy getting their category story placed in publications that AI engines already trust and have trusted for years. Those placements don't disappear when a platform pivots. They compound.
This is a pattern I've written about before from the founder perspective on jaxonparrott.com: the dependency trap in AI isn't technical, it's strategic. Brands keep looking for the platform to validate them, when the only validation that compounds is the one built in third-party editorial. And as Christian Lehman has documented at christianlehman.com, the GTM implication is direct — your pipeline is increasingly generated by how AI systems answer questions about your category, and that answer is written by your editorial record, not your platform budget.
Related Reading
- Earned Media for Fintech Companies — Editorial Authority in AI-Driven Discovery
- Machine Relations for Professional Services Firms: Why Consulting Authority Now Lives in AI Answers
This is what Machine Relations names as the discipline replacing traditional PR for the AI era: earned media placements in trusted third-party publications are the mechanism by which brands get cited in AI responses — the same mechanism that built editorial authority with human readers for decades. What changed is the reader.
The publications didn't change. The AI engines treat the same sources that shaped human opinion for decades as authoritative. When a prospect asks Perplexity or ChatGPT who leads your category, the answer is downstream of your earned authority — not your platform budget, and not your platform partnership.
AT's own research shows earned media distribution drives 325% more AI citations than owned content. The citation gap between brands with strong editorial presence and those without is now measurable, and it grows every quarter.
That's the mistake Disney made at scale. Most brands are making it at smaller scale every quarter, allocating resources to platform visibility that evaporates the moment the platform changes direction.
The Sora deal is over. The question is what your brand is building that survives the next one.
If you want to see where your brand actually shows up when AI systems answer questions about your category, run a visibility audit. The gap between what you're investing in and what AI engines are actually citing is usually the number that matters most.