‘This is not about short-term playtests’ — Disney’s AI-powered Star Wars Anzellans offer a glimpse at the future of its parks
Share this article
Join the conversation
Picture this: You’re walking through Black Spire Outpost on Batuu, the setting for Star Wars: Galaxy’s Edge at Disneyland or Disney World, and you step into a shop, maybe looking for a certain color of kyber crystal, a cold brew black caf, or a new piece of Star Wars merchandise. But before you can find what you’re after, you’re stopped by a small creature that starts talking to you.
Maybe it lets out a shriek of concern because you’re wearing an outfit that screams ‘dark side’, or cheers with excitement after spotting your birthday button. These are the kinds of spontaneous, memorable interactions Disney parks are known for, and in recent years, we’ve seen Imagineering steadily bring robotics and high-tech Audio-Animatronics closer together to make its characters feel more alive.
Now, Disney’s latest creation takes that idea a step further.
Latest Videos FromTechRadar
Meet the Anzellans: Vilo Vele, Kado Bre’a, and Jibu Bre’a. These tiny Star Wars creatures, the same species as Babu Frik from The Rise of Skywalker, have been brought to life as three Audio-Animatronics positioned together in a crate. They can move, look around, and interact with one another, but what’s especially impressive is that they can also read the room.
Thanks to a partnership with Amazon Web Services (AWS), the Anzellans use cameras, microphones, sensors, and AI to understand what’s happening around them and respond in real time. They can recognize what a guest is wearing, identify who’s speaking to them, and even coordinate their interactions with one another.

These Anzellans now are our most advanced Audio-Animatronics in the world
Kyle Laughlin
And while AI is helping bring these characters to life, it’s not writing their dialogue or dreaming up their personalities. That’s still very much the work of Disney’s Imagineers.
“These Anzellans now are our most advanced Audio-Animatronics in the world,” Kyle Laughlin, Senior Vice President of Walt Disney Imagineering R&D and Technology and Engineering, told me in an interview alongside Samira Panah Bakhtiar, General Manager of Telco, Media & Entertainment, Games, Advertising, and Sports at AWS.
Sign up for breaking news, reviews, opinion, top tech deals, and more.
It’s a pretty significant statement considering Disney’s long history with Audio-Animatronics, dating back to Walt Disney’s Enchanted Tiki Room, and continuing through increasingly sophisticated figures in attractions like Pirates of the Caribbean.
But the Anzellans are doing something different. Rather than performing a fixed sequence of movements and dialogue, they’re designed to understand what’s happening around them and decide how to respond, without someone controlling them from behind the scenes.
And it’s part of a much bigger effort underway at Imagineering.
Bringing robotics and Audio-Animatronics together

I’ve spoken with Laughlin previously about Disney’s characters, including the BDX droids that have been roaming Galaxy’s Edge and Olaf, the free-roaming robotic snowman that Imagineering has brought to Disneyland Paris and Hong Kong Disneyland.
While these creations might look very different from the traditional Audio-Animatronics that populate Disney’s attractions, they’re increasingly connected behind the scenes.
“Our robots and our animatronics are coming closer and closer together,” Laughlin explained.
In fact, the Anzellans are built on the same robotic character platform that powers Olaf and the BDX droids. That’s an especially interesting detail, considering these three little characters aren’t roaming around on their own. Instead, they’re working together in one environment, responding to guests and each other. Notably, there isn’t another Cast Member controlling the experience or puppeteering, but rather those decisions are made by the character, in real time.
This is our most intelligent character that we’ve created
Kyle Laughlin
Laughlin describes the Anzellans as a hybrid, bringing together elements of robotics and Audio-Animatronics with a level of intelligence that Disney hasn’t previously delivered in one of these characters.
“This is our most intelligent character that we’ve created,” he said.
The hardware is an important part of that distinction. While Olaf can move around and interact with his surroundings, Laughlin explained that he doesn’t have onboard cameras or microphones. The Anzellans do, along with additional sensors that help them understand their environment.
Imagine approaching the crate with a group of friends. The Anzellans can understand that multiple people are standing in front of them, determine who’s speaking, and look that person directly in the eyes while responding in their own mix of English and Anzellan.
It’s the kind of detail that could make an encounter feel much more personal, and it’s where AWS enters the picture.
AWS gives the Anzellans their intelligence, but Disney remains in control

Disney and AWS have worked together for more than a decade, including on the infrastructure supporting Disney+, which turns seven this week. But the Anzellans take that partnership in a different direction, bringing AWS’s cloud and AI capabilities into a physical character experience.
And the project came together surprisingly quickly.
Laughlin told me the teams worked together over a matter of months, with the opportunity to collaborate with the filmmakers behind The Mandalorian & Grogu helping inform the direction of the characters.
AWS I think brought the intelligence, Imagineering brought the brawn,
Samira Panah Bakhtiar
AWS brought its AI and software expertise, while Imagineering concentrated on the physical characters, their personalities, and the storytelling.
“AWS I think brought the intelligence, Imagineering brought the brawn,” Bakhtiar told me.
The Anzellans use Amazon Bedrock, AWS’s managed AI service, to access different AI models that help them understand and react to the world around them. Laughlin confirmed that Imagineering is using vision models, large language models, and world models as part of the experience.
That doesn’t mean the characters are using generative AI to make up their own dialogue, though.
Every line is written by Imagineers and prerecorded, and every character’s personality and movements are designed by the creative team. AI selects the appropriate responses and animations based on what’s happening in the moment.
So, if an Anzellan spots a birthday button, it can choose a response that fits the occasion. If someone speaks directly to one of the characters, it can respond accordingly. The interaction might feel spontaneous, but the actual words and character behaviors remain firmly in Imagineering’s hands.
That distinction is important, particularly as Disney continues to experiment with AI in guest experiences.
“Human creativity leads and the guest experience is preserved, and AI and the cloud computing infrastructure really is there to support it,” Bakhtiar said. “[Entertainment companies] need technology that’s going to really serve their story and not the other way around.”
There’s also another layer to this experience: the three Anzellans aren’t simply responding independently.

Bakhtiar explained that each character is simultaneously tracking three things: the guest, what’s happening in the surrounding environment, and what the other Anzellans are doing.
“The system is working like a show director, and it’s like keeping this ensemble together so the characters, they don’t talk over each other, the story stays coherent,” she explained.
That’s a fascinating distinction from a traditional Audio-Animatronics show. Instead of following the same carefully timed sequence, the Anzellans can adjust their performances to the situation while still behaving like characters from the Star Wars universe.
AWS also used Kiro, its AI coding agent, to help turn Imagineering’s creative briefs into working software. Bakhtiar explained that the teams reviewed and approved the work at each stage, allowing ideas to move into testing more quickly.
For Laughlin, that meant Imagineers could spend more time focusing on the mechanical and electrical engineering, character skins, and the finer details that make the Anzellans believable.
And because Bedrock isn’t tied to one specific AI model, Disney can swap models as the technology improves, optimizing for performance, speed, latency, or cost without rebuilding the entire experience.
“We can be on the bleeding edge of whatever model is launched in order to bring the best experience possible,” Laughlin said.
That’s particularly valuable when a character needs to respond quickly enough for a conversation to feel natural.
A new kind of character encounter

Of course, there’s a big question hanging over all of this: When will guests actually get to meet the Anzellans?
Disney isn’t saying just yet. The characters are scheduled to appear at AWS re:Invent in Las Vegas from November 30 through December 4, where attendees can meet them at the Customer Showcase at Caesars Forum.
And while this won’t be their first public appearance — Disney confirmed to TechRadar that an earlier version of the Anzellans shown at D23 was also powered by AWS — there’s still no word on when or where they’ll become a regular guest experience at Disneyland or Disney World.
Laughlin did offer one encouraging indication when I asked about their eventual arrival in the parks.
“We’re excited for guests to meet them. We just won’t talk specifics today.”
And while we’ve already seen the BDX droids roaming Galaxy’s Edge and Olaf interacting with guests at Disneyland Paris, these experiences are still relatively limited compared with the many Audio-Animatronics and characters found throughout Disney’s parks. The technology is impressive, but the next challenge is bringing these more interactive characters into everyday park experiences, and doing so at scale.
That’s something Laughlin and I have discussed before, and he sees the Anzellans as part of a broader shift in how Imagineering develops and introduces its characters.

“More than ever, this is not about short-term playtests. It really is about populating our worlds with lots of different characters,” he explained.
Laughlin also told me that guests should expect to encounter the Anzellans indoors first, and his description of the technology offers a pretty intriguing glimpse at where Imagineering might be headed.
“We’ve viewed this as an experiential merchandise platform,” he explained.
In other words, imagine walking into a Disney shop and having an unexpected interaction with a character, rather than needing to visit a designated meet-and-greet location.
And the idea isn’t necessarily limited to Star Wars.
“When we think about the future of where it’s headed, you can imagine that any type of merchandise experience could be powered with characters from any of our franchises,” Laughlin said.
There’s still plenty to figure out, though. Keeping a character responsive throughout a busy day at a theme park requires more than just sophisticated AI. There’s also the question of what happens if the network connection drops or the primary system encounters a problem.

It’s something I asked Laughlin about, and he explained that Disney is combining AWS cloud and edge infrastructure to help keep the experience running.
“If something were to fail […] then we could always fall back to another experience,” he said.
The goal is to keep these characters operating reliably throughout park hours, though Disney is still learning what it takes to deliver these more intelligent experiences at scale.
Privacy is another consideration, especially with cameras and microphones involved. Laughlin told me that the information gathered during an interaction is used only for the immediate experience, isn’t currently stored in the cloud for Disney to reuse, and isn’t used to train AI models.
The specifics of the underlying models and the number of sensors remain undisclosed, but the broader approach is becoming clearer.
The next chapter for Disney’s characters

Imagineering has continued to push its characters further, from free-roaming robots to the next-generation Audio-Animatronics appearing in attractions like Pirates of the Caribbean or Frozen Ever After.
The Anzellans are another step in that evolution, but perhaps the bigger development is what’s happening behind the scenes. Disney is building a shared foundation that can support different types of characters and experiences, rather than treating robotics, Audio-Animatronics, and AI as entirely separate technologies.
Bakhtiar described the AWS-powered system as a reusable framework, one that could allow Imagineers to focus on creating new personalities, moods, movements, and stories instead of rebuilding the underlying software every time.
And with Laughlin pointing to more than $60 billion of work ahead for Imagineering worldwide, there are plenty of opportunities to put these technologies to work.
Just look at Walt Disney World, where an entirely new Monsters, Inc. land is coming to Hollywood Studios — complete with an octopus-like, impressive animatronic — while Animal Kingdom is getting Tropical Americas with new Indiana Jones and Encanto attractions. Over at Magic Kingdom, Villains Land and Piston Peak are on the way, while EPCOT will eventually take guests on a new journey into imagination with Figment and Dreamfinder.
And that’s just Florida. At Disneyland Resort in California, Imagineering is working on the first-ever Coco attraction, a major reimagining of Tomorrowland, and an expansion of Avengers Campus, among other projects.
Of course, that doesn’t mean we’ll necessarily see the Anzellans or this exact AI technology in any of these attractions. But with so much happening across Disney Parks, it’s easy to see why Imagineering is investing in a platform that could bring more interactive characters to life across its parks.
my hope is over the next 10 years you’re going to continue to see an incredible pace of scale and delivering these experiences worldwide
Kyle Laughlin
For now, the Anzellans remain a glimpse at what’s possible rather than a confirmed new fixture at Galaxy’s Edge. But it’s easy to imagine how a seemingly ordinary shopping trip could become something much more memorable when a character recognizes you, reacts to what you’re wearing, or strikes up a conversation.
And when I asked Laughlin what this convergence of robotics, AI, and Audio-Animatronics could mean for Disney’s characters over the next decade, he pointed to the growing range of possibilities, from robots and Audio-Animatronics to puppets and roaming characters.
“We now have more tools in the toolkit to bring these characters to life in all kinds of new ways,” he said. “And so, my hope is over the next 10 years you’re going to continue to see an incredible pace of scale and delivering these experiences worldwide.”

Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!
Source
Share this article Join the conversation Follow us Add us as a preferred source on Google Newsletter Subscribe to our newsletter Picture this: You’re walking through Black Spire Outpost on Batuu, the setting for Star Wars: Galaxy’s Edge at Disneyland or Disney World, and you step into a shop, maybe…
Recent Posts
- ‘People will ask, agents will act’: Workday CEO outlines how AI is coming to help unleash “an incredible world of work”
- The Verge’s 2026 holiday gift guide for travelers
- 14 Elite Coffee Gifts for Caffeinated Connoisseurs
- How The Verge is picking which Black Friday deals to share
- ‘This is not about short-term playtests’ — Disney’s AI-powered Star Wars Anzellans offer a glimpse at the future of its parks
Archives
- November 2026
- October 2026
- September 2026
- August 2026
- July 2026
- June 2026
- May 2026
- April 2026
- March 2026
- February 2026
- January 2026
- December 2025
- November 2025
- October 2025
- September 2025
- August 2025
- July 2025
- June 2025
- May 2025
- April 2025
- March 2025
- February 2025
- January 2025
- December 2024
- November 2024
- October 2024
- September 2024
- August 2024
- July 2024
- June 2024
- May 2024
- April 2024
- March 2024
- February 2024
- January 2024
- December 2023
- November 2023