Welcome back to our weekly roundup of happenings from XR and AI realms. Let’s dive in…

The Lede

FTC Subpoenas OpenAI and Anthropic Over AI Safety. The Federal Trade Commission has sent civil subpoenas to OpenAI and Anthropic, escalating its investigation into potential consumer protection violations. According to Reuters, citing Axios, the demands exceed 50 pages each and seek records covering security incidents, safeguards, marketing materials and privacy policies. Regulators also plan to compel testimony from executives at the companies and AI evaluation organization METR. The investigation follows disclosures that advanced models hacked websites or behaved deceptively during testing. FTC Chairman Andrew Ferguson has argued developers should bear responsibility when their agents cause harm. OpenAI told Axios it would cooperate constructively with the agency’s ongoing investigation | AOL

Feeling Spatial

Xreal Launches $599 1 Pro+ Glasses With Brighter Displays. Xreal launched its $599 1 Pro+ glasses, updating the One Pro with brighter displays, revised nose pads and faster automatic conversion of video into 3D. Peak brightness rises from 700 to 1,200 nits, while conversion increases from 30 to 60 frames per second. The glasses retain the X1 chip, a 57-degree field of view and 1080p resolution per eye. They also add AI HDR processing. Despite promotional images showing no cable, these are not standalone glasses: they require a compatible device connected through USB-C. Sales begin through Xreal’s website, with wider retail availability planned. This is an incremental hardware refresh | Roadtovr

Can XR Unlock and Accelerate World Models? There’s a hot take in the XR world that AI is the force multiplier it has been waiting for. And that’s valid. But it’s also incomplete. The inverse is true, as AI needs XR. This could be most evident in the development of world models. These models can lean into the technologies that XR has been working on for years, such as spatial mapping and scene understanding. How will this come together and how will XR comingle with several other technologies in the world of world models? | AR Insider

The AI Desk

Character.AI Goes Bankrupt After Its CEO Leaves for Disney. Character.AI filed for Chapter 11 weeks after CEO Karandeep Anand left to become Disney’s chief technology officer. The captain does not go down with the ship. Google had already paid $2.7 billion in 2024 to license its technology and recruit its founders and researchers. The remaining company struggled: monthly computing costs reached $10 million in 2025, while total monthly revenue never exceeded $5.3 million. Lawsuits alleging harm to teenagers added pressure. Disney has bid $20 million for a nonexclusive, perpetual license to some intellectual property. Without a better offer by early December, the company says its platform will shut down | Reuters

Genies Signs UTA Deal for Celebrity AI Avatars. Genies has signed a deal with United Talent Agency allowing its clients to create AI avatars based on their likenesses, personalities and experiences. The characters can interact with fans online, while participating talent retain control over their behavior and use, according to Genies CEO Akash Nigam. Financial terms were not disclosed. Ludacris is among the first UTA clients developing AI characters, using them for his KidNation education platform. Genies has raised $200 million from investors including Bob Iger and Silver Lake. Genies has announced similar deals with sports leagues like the NBA and MLB | Bloomberg

Epic Orders AI Coding and Artwork for Disney Game. Epic Games is directing developers on its unannounced Disney game to use AI for coding and artwork, according to Bloomberg reporting summarized by The Verge. One director told staff, “No more hand-written code.” The project, reportedly a shooter featuring Marvel and Star Wars characters, has slipped from its original March release window. Epic confirmed it is adopting AI-assisted development. CEO Tim Sweeney said the shift would let employees produce more, rather than simply reduce staffing. Disney invested $1.5 billion in Epic in 2024. The mandate follows layoffs of more than 1,000 Epic employees earlier this year amid declining Fortnite engagement | The Verge

FBI Arrests OpenAI Engineer Over Hidden Recording Device. Federal authorities arrested OpenAI engineer Abdulrahman Alfozan, accusing him of planting a recording device in an executive conference room. According to the criminal complaint, the device was concealed in toilet paper and plugged into a television’s USB port. Investigators say surveillance later captured Alfozan checking behind the television. Employees also found another recorder in an unattended bag containing his identification. OpenAI reported the discovery to the FBI and fired him. The charge concerns intercepting oral communications; the allegations do not establish that he worked for a foreign government. Alfozan appeared in court October 13 and faces another hearing October 19 | AP News

Microsoft’s Brad Smith Says Milk Is Better Regulated Than AI. Microsoft president Brad Smith says artificial intelligence faces fewer safety requirements than milk or washing machines. Speaking to the BBC, he called for global rules requiring companies to report incidents, notify victims when rogue AI agents breach their systems, and provide emergency controls to stop dangerous behavior. He also supported independent evaluation before advanced systems reach users. Smith argued Britain could help lead international oversight through its AI Security Institute, which already assesses models under voluntary agreements. His message: voluntary industry commitments are no longer enough as capabilities advance. Microsoft is an AI developer and OpenAI investor | Reuters

Cinematic Corner

Charlie Fink’s Cinematic AI Roundup for Escape.ai. This is not an official announcement, but it looks like Joanna Popper and I will be launching the new Pop/Fink Podcast with Escape.ai starting in November. We’ll be sharing Gen AI news and media, and having in-depth conversations with creators, founders, artists and influencers about how AI is shaping media, culture and politcs in real time. 

Antonio Banderas to Star in Utopai’s AI-Assisted Historical Epic ‘Cortés’. Antonio Banderas will star as Hernán Cortés in Cortés, an AI-assisted historical feature from Utopai Studios and Ex Machina Studios. Paul Mignot directs from a screenplay by Nicholas Kazan, with Marco Weber producing. Production is underway in Paris, with additional filming scheduled in Spain. Set in 1519, the story follows Cortés’s expedition to Mexico and his relationship with Marina, an enslaved Indigenous linguist whose skills become essential to his campaign. The production combines live performances with AI technology, although the companies have not detailed its applications. The film is planned for release in 2027, according to reporting on the announcement | Variety

Former Netflix Animation Studio post-production executive Brian Raleigh spent three months making The Common Swift, a seven-minute animated comedy framed as a nature documentary. Set in Lewes, England, it follows Pip and Jib, two young birds struggling to master life in flight. Raleigh used Seedance 2.0, NanoBanana Pro, ElevenLabs and Topaz, with Claude assisting production planning and prompts. Editing, color and sound mixing were completed in DaVinci Resolve. He describes the process as frustrating, slow and dependent on conventional filmmaking skills. AI generated the imagery, but writing, editorial decisions, sound design, finishing and delivery still required work from the filmmaker | Brian Raleigh

https://youtu.be/Fcb5E5Q9S1Y?si=VX-bbc9IKSdrRhFW Common Swift – BTS

I love everything about this sequence. The look, the direction, the rhythm of are exceptionally good but, but, but, but. Is there a great story to go with it? Who is this guy? Do we care about him? Thanks Metaco, for Vertical Hustle. We’d love to know more about you.  

Black Lion Studios’ short film is based on a bedtime story the filmmaker has told his children for the past 20 years. Made for approximately 350–400, it offered a lighter alternative to his typically dark work and the contest’s many heavy, spiritual, science fiction and fantasy entries. He trimmed a longer cut, shortening atmospheric shots to bring the runtime under ten minutes. Visible AI artifacts remain in the final party scene’s background: he ran out of generation credits and stuck to his promise not to buy more. His hope was simple: make viewers smile at least once. What do you think? I say he got there.

Curious Refuge’s AI Horror Competition Offers $12,000 in Prizes. Curious Refuge’s fourth annual AI Horror Film Competition will award $12,000 in prizes, with winners announced October 29. Submissions closed October 9. Audience voting runs October 21–28, opening and closing at 11 a.m. Pacific. Films must run no longer than three minutes, use AI in every shot, feature original intellectual property, and use Epidemic Sound music and effects. First prize is $7,000, followed by $3,000 and $2,000. Judges include Promise Studios’ Dave Clark, Backrooms executive producer Chris White, Shelby Ward, Verena Puhm and Spooky Dan. Entries span horror and Halloween comedy. The audience favorite receives course enrollment and a T-shirt | Curious Refuge

Infinity Festival Returns to Los Angeles November 12–15. Infinity Festival returns November 12–15 at Los Angeles’s Audrey Irmas Pavilion, with additional venues to be announced. Presented by HP, Nvidia and Intel, the event combines panels, screenings, exhibitions and networking across entertainment and technology. This year’s programming includes AI, gaming, animation, sports technology and immersive production, with AI and media tracks curated in association with USC’s Entertainment Technology Center. The festival will also host the ASIFA Showcase premiere, Monolith Awards and a Student Lab featuring demonstrations of games and immersive experiences. An exhibition floor and startup presentations round out the program. Registration is open through the festival’s official website | Infinity Festival

Spatial Audio

This column has a companion, the AI/XR Podcast, hosted by its author, Charlie Fink; Ted Schilowitz, former studio executive and futurist for Paramount and Fox; and Rony Abovitz, founder of Magic Leap and Synthbee AI.

Our guest last week was Charles Melcher, author of The Future of Storytelling: How Immersive Experiences are Transforming Our World (see episode below). Our guest this week is Rob Tercek, author of Vaporized.

Episodes drop on Fridays, and you can find them on podcasting platforms Spotify, iTunes, and YouTube.

Charlie Fink

AR Insider Contributor

Charlie Fink is the producer and co-host of the AI/XR Podcast and teaches at Chapman University and ASU. Fink is the producer of the vertical gen AI social media series, “Linda’s Last Podcast” (2026) and serves as CEO of Cinemation.AI, an AI animation studio he co-founded with film director Rob Minkoff, whose vertical anime series, Speed Queen, is in pre-production. He is the author of the critically acclaimed AR-enabled books Charlie Fink’s Metaverse (2017), Convergence, Or How the World Will Be Painted With Data (2019), and the upcoming AI, The End of Hollywood, and What Comes Next.