Hollywood Legend Turned Startup Investor Jeffrey Katzenberg On Where He’s Placing His AI Bets

When Jeffrey Katzenberg sold DreamWorks Animation in 2016, he knew his next venture would be technology related. Soon after the sale, he teamed up with former Dropbox CFO Sujay Jaswa to found WndrCo, a holding company and venture capital firm “reimaginging how people live and work.”

With $2.8 billion in assets under management, the firm has invested in companies including Writer AI, 1Password, Abridge, Airtable, Alembic, Aura, Cursor, Databricks, Figma, Point Wild and Super Unlimited.

Film producer to tech investor may not seem like the most natural of evolutions. But for Katzenberg, it wasn’t a stretch.

Before DreamWorks, he was chair of Walt Disney Studios, which during his tenure, produced films such as “Aladdin,” “The Lion King,” and “Beauty and the Beast.” He was also previously president of Paramount Pictures.

Over the course of his illustrious and legendary career, Katzenberg experienced firsthand how technology could impact the film industry — from Steven Spielberg building a big shark out of plastic and dragging it behind a motorboat as a “special effect” to the “transformative” power of animation.

“I always saw tech as a competitive advantage in storytelling,” Katzenberg told Crunchbase News in an interview.

The introduction of computer graphics and CGI didn’t evolve animation, he said, but “revolutionized” it. And that, he said, is somewhat analogous to what’s going on with AI today, and not just in media, but “writ large.”

“It’s actually a fantastic, interesting use case of where you can see not only total and complete reinvention and disruption, but the ultimate outcome of something bigger and better,” Katzenberg said. “It displaced every job, but it did not eliminate jobs. And that’s where I actually have a slightly more optimistic point of view about what is coming our way [when it comes to AI].”

Today, Katzenberg and WndrCo general partner Justin Wexler lead the firm’s enterprise AI investing practice and work with portfolio companies to secure big customer wins, helping them land deals with brands like Nike, Delta, NBA and Disney.

WndroCo Holdco, the firm’s company-building arm, has incubated eight companies with a total of a combined $1.25 billion in top-line revenue and $250 million in EBITDA. Through this strategy, WndrCo partners with founders and CEOs, often acquiring controlling stakes in “underappreciated” tech companies in an effort to turn them into category leaders.

Crunchbase News conducted a video interview with the pair to get their thoughts on what’s next when it comes to enterprise AI and how storytelling can be a competitive advantage.

The interview has been edited for clarity and brevity.

Where are we exactly in the enterprise AI cycle?

Katzenberg: We are of the school that the world is being revolutionized, not “evolutionized,” by the introduction of AI on the third anniversary of ChatGPT and all of the amazing things that have happened since then.

There are many ways to look at AI. It’s such a giant umbrella … and it has so many verticals in it … and so many layers, some of which are out of our realm. We’re not in the hyperscaler business. We’re not in the LLM business. We’re not in the infrastructure side of it. … We are most excited about the simplest applications that actually change how people go about their lives, be it at work or at home in their professional lives or on the consumer side of it.

The overarching category would be the consumerization of software. So we’re trying to stay very focused on this idea that if you actually just took what has already been created with these large language models, and where we are in January 2026, the value that can be built today with the tech that has already been created and is deployable, could be bigger than everything that has come before it.

Wexler: We’re in a pretty unique role in this ecosystem that we have a lot of pride in. Jeffrey has built so many relationships over decades, and a lot of folks admire the transformation he had to go through in animation. So we just try to be a really close partner to a lot of Fortune 500 execs, as they’re thinking about, “There’s this totally game-changing technology with generative LM AI and agentic AI. How do I deploy it to transform my part of the organization?”

…Getting deployments at scale that are successful is hard to do, and it’s hard for many reasons. There’s a lot of confusion in a lot of these organizations on how best to set up AI for a lot of success. It’s hard for a startup alone to try to be learning about an organization while selling and while competing with a lot of other noise in the market. … One thing that we spend a lot of time with the founders we work with is helping them navigate that … because if you can prove success, then you can really scale. It’s also a huge unlock for enterprise. But to date, there’s been a lot of deployments that, because you don’t have that alignment, it’s a lose-lose for everybody.

We actually counted one company that did like 200 meetings with a startup, and it went nowhere because the alignment wasn’t there. The technology could have been very impactful to them, but they didn’t have that organizational alignment. So that ingredient is so important. We’ve tried to play that role.

Katzenberg: I think that’s where our role is becoming more important and valuable, and that can be summed up in a simple word: trust. We’ve now been doing this long enough that we have built out a pretty extraordinary network of relationships with the Fortune 500 companies and the C-suites there, and we have brought multiple products to them over the past couple of years, each one of which has actually delivered on its promise.

So today, I find that when we come in with a new product, a new offering, a new company, it’s easier; it’s not easy, but it’s easier.

We are super sensitive around the issues of reliability, security, data protections and all of the things that we already know we can anticipate where people are going to be anxious.

You have said that you work closely with your companies and landing these big logos and clients. What separates a cool AI demo from a product that actually does become embedded in enterprise workflows?

Wexler: We found that it’s hard to build for everybody all at once. There’s incredible AI and technology that serves SMB and mid-markets. But the founders we’ve seen that have the most success in enterprise, built for enterprise from day one. They didn’t try to move up, as they’re getting some traction with startups. They built with compliance from day one. That way, it’s much easier to get HIPAA compliant and GDPR versus trying to make changes along the product journey.

… so you can scope a great initial proof of concept or pilot, but they’re betting just as much on this potentially being a new partner for them for the longer term. We’ve seen a one-year deal turn into a three-year deal, and actually a couple companies are starting to talk about four- or even five-year contracts, which we have not seen previously with AI startups. But it’s happening now, and just goes to show the level of comfort that enterprises are gaining, and the trust that they’re gaining with some of the most innovative early-stage companies that could be long-term partners for them.

When you look at AI startups today, how do you decide whether something is actually an AI company or really just like a feature on top of somebody else’s model?

Wexler: We have a portfolio of companies that we would categorize as AI, and they partner very well with the major LM companies, whether it’s OpenAI, Anthropic or Gemini. We also have companies like Writer, which have trained off of the models either from scratch or from working Llama models as well, where they really take an ownership in trying to build a best-of-breed model for their specific use case. There’s been success with both.

Where do you expect most of the enterprise value to actually accrue? Would it be an infrastructure foundation? Models, application layer?

Katzenberg: I think 100% at the application level, not 80% or 90%, but 100%, as best we are seeing. That’s where you can actually see incremental value. It is additive and it is not, rip and replace. Now, will this accelerate in coming years? Likely. But this idea that in the next five years, 80% of white-collar workers are going to be replaced, I think, is just way ahead. I’m not saying that might not actually be correct in the long run, but the transition to it is going to be, I think, significantly more than five years.

Wexler: As application-tier companies build their case studies and prove examples, the rest of those industries pay close attention, see the success and want to follow. So then as application-tier companies scale, they’re actually the ones who choose the infrastructure that’s best for them, and that includes the LLMs as well. Really, what the end client, the Fortune 500 or the enterprise, is getting is kind of a package solution. A lot of people were trying to build CRMs and ERP before Salesforce 1 really built the bundled solution that became the industry standard. We’re going to see that for particular use cases. Look no further than Harvey or Bridge.

Katzenberg: The big hyperscalers probably have the capability of doing it all. But they don’t have the capacity, and they don’t have the focus. And so that’s why, when you look at a Bridge, or you look at Harvey or Cursor, in these cases these are phenomenal entrepreneurs who are brilliant and singularly focused. They have this one idea, this vision if you will, for an exceptional application with a value.

So can somebody come along and do what any of those three companies can? They can’t do it all, and they’re not going to do it with that same laser ambition and focus of a great entrepreneur. They might have the capabilities to do it, but I don’t think they have the focus. They can’t. When you try to be all things to all people, you end up not being anything.

Looking ahead to 2026, what do you expect the biggest correction to be — whether in valuations, business models or assumptions founders and investors are making about enterprise AI today?

Katzenberg: I have a bottomless well of optimism. I’m not looking for the clouds on the horizon. Where’s the concern? Where are the anxieties coming from? I’m not saying that it’s not there. It’s just in our lane, there’s too much opportunity I do feel we are well equipped to navigate.

In AI specifically, what are some nonnegotiable qualities that you look for in a founder?

Wexler: It’s a pretty high bar to be a great AI founder. There’s a lot of demands on you. You’re at the cutting edge of technology where things are constantly shifting faster than any other era. You have to stay on top of that and be on the bleeding edge. Because if not, it’s going to be hard. You have to be able to truly understand these pain points in a way that you can articulate how technology can solve them, and in a way that has not been solved before.

How does great storytelling, in your opinion, actually change outcomes?

Katzenberg: One of the things that was the siren drawing me to Silicon Valley and company building is I realized very, very early on how foundational storytelling is to every single chapter of every single company’s journey in that the better you can tell your story, the more successful you are.

Telling your story in every phase of company building is more than essential, it’s existential. The way in which you get other people to join you and pursue your idea as an entrepreneur, you’ve got to be able to share your dream in a way that others can get excited about, and recognize your vision, your ambition, your dream.

You have to be able to tell it to get other people to come on the journey with you. You need to tell it to investors in order to be able to raise money. You need to be able to tell it to customers when you go to market, and virtually every step along the way in company building and company running.

Related reading:

Illustration: Dom Guzman