OpenAI adds reusable ‘characters’ and video stitching to Sora

OpenAI's Sora Introduces Character Cameos and Video Stitching in Major Feature Update

OpenAI has unveiled significant new capabilities for its Sora AI video platform, introducing reusable character avatars and multi-video stitching functionality. The updates mark a pivotal shift from single-clip generation to comprehensive storytelling tools, positioning Sora as a more versatile creative platform.

The new "character cameos" feature allows users to transform virtually any image into a persistent digital avatar. Whether it's a beloved pet, custom illustration, or favourite toy, these characters can now appear consistently across multiple video projects.

Digital Personas Get Granular Privacy Controls

The character creation system offers three distinct sharing levels: private use, mutual follower access, or full community availability. Each character receives its own display name and handle, making it easy to tag into future video projects.

"Once created, each character comes with its own permissions, separate from your personal likeness: keep it just for yourself, share it with mutual followers, or open it to everyone on Sora," according to OpenAI's release documentation.

This granular control addresses growing concerns about digital identity management in AI-generated content. Users can experiment with creative personas while maintaining strict boundaries around personal likenesses.

For creators looking to explore these possibilities, our beginner's guide to using Sora AI video provides comprehensive getting-started instructions.

Legal Turbulence Around "Cameo" Terminology

The feature's naming has sparked immediate legal challenges. Celebrity video platform Cameo has filed a trademark infringement lawsuit against OpenAI, objecting to the use of "cameo" in Sora's feature descriptions.

This legal dispute mirrors broader intellectual property tensions in the AI space. Similar conflicts have emerged across the industry, as we explored in our analysis of Warner Bros taking Midjourney to court over AI and superheroes.

For related analysis, see: The Skeleton Key AI Jailbreak Technique Unveiled.

The case raises questions about how AI companies can navigate existing trademark landscapes whilst developing new product categories. It also highlights the speed at which legal challenges materialise in fast-moving tech sectors.

Multi-Scene Storytelling Through Video Stitching

Beyond character management, Sora now supports video stitching capabilities that enable creators to connect multiple scenes into cohesive narratives. This functionality transforms the platform from a single-clip generator into a comprehensive storytelling toolkit.

The stitching feature works seamlessly with the new character system, allowing consistent personas to appear across different scenes within longer compositions. Users can now craft multi-act stories, tutorials, or complex promotional content.

"We are launching the ability to buy extra generations in Sora today... there simply isn't enough computing power currently available to sustain unlimited unrestricted use," explains Bill Peebles, OpenAI's head of Sora.

For related analysis, see: OpenAI AI Pen May Light Up Note-Taking in 2026.

Community Engagement and Competitive Dynamics

OpenAI has introduced leaderboards highlighting the most remixed videos and frequently used characters. This gamification approach encourages community participation whilst showcasing popular creative approaches.

The timing coincides with increased competition in AI video generation. Platforms like Kling, the Chinese AI video model outshining Sora, continue gaining market traction through different technical approaches.

To boost accessibility, OpenAI has temporarily removed invitation requirements for users in the US, Canada, the UAE, and Korea. This "limited time" open signup represents a significant shift from the platform's previously exclusive access model.

For related analysis, see: The MENA AI Startup Ecosystem - Accelerating Growth and Inno.

Key community features include:

• Most remixed videos leaderboard tracking viral content patterns
• Popular character rankings showing community preferences
• Enhanced sharing controls for collaborative projects
• Cross-platform availability spanning iOS and Android ecosystems

Despite initial enthusiasm, Sora faces challenges maintaining user engagement. The platform's download numbers have declined 45% month-over-month, suggesting initial curiosity may be giving way to more measured adoption patterns.

Can I upload characters created with other AI tools?

• OpenAI's documentation mentions "original personas" generated within Sora, but it's unclear whether externally created AI characters are supported. The policy likely aims to maintain quality control and avoid copyright complications.

How realistic can character cameos be?

• The photorealism limits remain undefined in current documentation. This ambiguity raises questions about distinguishing AI-generated persons from real individuals, particularly regarding consent and identity verification protocols.

For related analysis, see: Robots Take Stage as Backup Dancers.

Will character cameos work with existing video projects?

• The feature appears designed for new video creation rather than retroactive application. Users will likely need to regenerate content to incorporate newly created character avatars into their projects.

How long can stitched videos become?

• OpenAI hasn't specified maximum lengths for stitched compositions. The limitation likely depends on computational resources and subscription tier, given the platform's recent move to paid generation models.

Are character cameos available globally?

• Character features should be available wherever Sora operates, but the temporary open signup currently covers only the US, Canada, the UAE, and Korea. Other regions still require invitation codes.

Further reading: OpenAI | Reuters | OECD AI Observatory