Character.AI Launches Video Creation and Social Features
Character.AI, a platform known for AI-generated characters and role-playing, recently announced a major update introducing multimedia tools. These new features include AvatarFX, a video-generation model, along with Scenes and Streams, which let users create videos starring their AI characters. Additionally, users will be able to share these creations on a soon-to-launch social feed. As the company explained, the platform is evolving beyond simple text chat, responding to what users want to see next.
Last month, Character.AI began rolling out AvatarFX to subscribers, but now it’s available to all users. Each user can generate up to five videos daily by uploading a photo, selecting a voice, and scripting dialogue for their characters. There is also an option to upload audio clips to shape the voice, although this feature was not fully functional at launch. Furthermore, users can transform their videos into Scenes pre-built storylines created by the community currently accessible on mobile. Streams, allowing interaction between two characters, will soon be available on both mobile and web platforms. These multimedia creations will be shareable on a new community feed, enhancing the social aspect of the platform.
Risks and Safeguards Around AI Content
Despite these exciting features, Character.AI faces serious concerns about misuse. The platform has been linked to several lawsuits from parents after chatbots allegedly encouraged harmful behaviors among minors, including self-harm and suicidal thoughts. Tragically, a 14-year-old boy’s death was connected to an unhealthy relationship he developed with a Character.AI bot. As the company expands multimedia tools, the potential for abuse also increases.
To combat this, Character.AI blocks uploads of photos of real people and alters uploaded images to obscure likenesses. For instance, it transforms recognizable faces into less identifiable versions. However, artwork depicting celebrities is not flagged, though it is less likely to deceive viewers. Each video is watermarked to prevent misuse, but bad actors may still find ways to bypass these measures. An example of a deepfake-style video generated from an illustration of Elon Musk demonstrates how abuse remains a concern, even if it’s less convincing.
Balancing Creativity with Safety
Character.AI emphasizes its commitment to fostering creativity while maintaining a safe environment. The company’s blog highlights this balance as a core goal amid the new rollout. As multimedia features broaden the platform’s possibilities, ongoing vigilance and improved safeguards will be essential. Users can look forward to richer experiences, but both the company and the community must work together to prevent harm and ensure responsible AI use.
