A TikTok Clone With an AI Core

The app will offer a vertical, swipe-to-scroll feed where users can watch and create AI-only videos up to 10 seconds long. Unlike TikTok or Instagram Reels, you won’t be able to upload clips from your camera roll or other apps. Instead, every video is created inside the platform using Sora 2’s text-to-video technology.

Users will reportedly be able to:

• Use their own likeness to inspire AI-generated videos.
• Tag other users, with notifications sent if someone’s appearance shows up, even in draft clips.

Think of it as a deepfake generator on steroids, though OpenAI says you’ll at least be notified if your face is used.

Employees Are Obsessed, Managers Are Worried

Wired reports that OpenAI launched the app internally last week, and early reactions from employees were “overwhelmingly positive.” Staff have reportedly been using it so much that some managers joked it was distracting from work productivity.

OpenAI hopes this app will do for AI video what ChatGPT did for AI text, helping the public understand and play with generative video tools in a fun, social way.

The Competitive AI Video Race

This move places OpenAI directly into an emerging AI-video social media war:

• Meta recently launched Vibes, an AI video feed inside its Meta AI app. So far, it’s been met with ridicule and scepticism online, with some critics calling it “pure garbage.”
• Google has integrated its Veo 3 text-to-video model into YouTube Shorts, aiming to make AI video a mainstream creator tool.
• OpenAI’s timing could benefit from the ongoing uncertainty around TikTok’s future in the U.S., potentially carving out a space for an alternative platform without ties to China.

If successful, OpenAI’s app could become the first widely adopted, AI-only social platform, something no other tech giant has pulled off yet.

Big Questions Remain

While the concept is intriguing, it also raises serious challenges:

• Misinformation & deepfakes: TikTok has already updated its community rules to label AI-generated content and ban harmful misinformation. OpenAI will likely face intense scrutiny to prevent fake news, scams, and malicious impersonations.
• Copyright issues: AI video generators have already been hit with lawsuits for training on copyrighted material and enabling unauthorised likeness use.
• Youth safety: OpenAI has faced criticism over how under-18s use ChatGPT; launching a highly visual, viral platform could magnify those concerns.

OpenAI hasn’t detailed how it will address these issues, but given the spotlight on AI safety and regulation, the company is expected to tread carefully.

Conclusion

OpenAI has not announced a public launch date for the app or for Sora 2, but the internal rollout suggests a release may not be far off. If the app lands soon, it could:

• Showcase Sora 2’s capabilities in a mainstream, user-friendly way.
• Push AI-generated short-form video into the social media mainstream.
• Position OpenAI as a direct challenger to TikTok, Instagram Reels, and YouTube Shorts, but with a feed made entirely by machines.

Whether this turns into a creative playground or a flood of AI slop nobody wants remains to be seen. But given the excitement from OpenAI’s own employees, and the race among tech giants to own AI video, this could be one of the boldest social media experiments of 2025.