Pika expands into persistent AI Selves

Pika is moving from a tool that makes clips into a product that can stand in for a person. The jump from prompt based video generation to AI Selves changes the job from making one output at a time to maintaining an ongoing presence across chat, social, and work apps. That opens new use cases like answering messages, posting content, and joining live video calls, which is a much broader surface than editing or generating short videos.1 2

• The current AI Self product is not just an avatar skin. The live site says users upload a selfie, record their voice, answer questions, and then connect that identity to Slack, Telegram, WhatsApp, X, Instagram, LinkedIn, YouTube, Notion, GitHub, Zoom, and more. That is closer to an agent plus digital twin than a video app.2
• This fits a wider shift in generative media, where value moves from raw model output toward workflow and automation. In adjacent research, OpenArt describes basic video tools as manual, where users stitch clips, audio, and edits themselves, then differentiates by automating more of the storytelling stack. Pika is pushing in a similar direction by expanding from single videos to persistent identity and cross app actions.3 1
• It also gives Pika a second distribution and monetization path. The core business still sells generation credits for clips and effects, and Pika also gained reach through Adobe Firefly Boards as a partner model inside a major creative suite. AI Selves adds a separate product wedge aimed at communication, creator presence, and lightweight work tasks, not just media generation.1 4

The next step is for Pika to bundle creation, identity, and automation into one consumer media stack. If the company can turn a selfie based video user into someone who also deploys an AI Self to post, reply, and create across apps, it stops being just another video model and becomes a broader personal media platform.1 2 4