OpenAI lets you control how your face appears in Sora 2

OpenAI’s Sora 2, its headline-grabbing AI video generator, has been the talk of the internet for how easily it can bring text prompts to life. One of its flashiest features has been the ability to add yourself, or anyone else (with consent), into AI-made videos. But that same power raised a big question. What happens when someone uses your likeness in a way you never agreed to?

After a wave of feedback, OpenAI is now giving users the power to control their digital cameos. Bill Peebles, Head of Sora at OpenAI, announced the update on X, explaining that users will soon be able to decide exactly how their likeness can appear in videos.

ALSO READ: OpenAI launches Sora app for AI-generated short videos

Through a new “Edit Cameo” section in Sora 2’s settings, users can set detailed preferences, like blocking political content, banning certain keywords, or limiting reuse altogether. Think of it as a digital boundary line for your AI self. OpenAI says this move aims to promote creative freedom without compromising user consent or safety. In simpler terms, you’re now the director and producer of your own AI identity.

Safer, clearer, and more transparent

Alongside the cameo controls, OpenAI has also introduced a new watermark system for Sora videos that makes AI-generated content easier to identify. The company says it’s now “more visible and clearer”, helping creators be transparent while ensuring audiences know when something was AI-assisted.

Behind the scenes, OpenAI has also upgraded Sora’s moderation tools to detect potential misuse more accurately and reduce false negatives, cases where inappropriate or harmful content slips through the cracks.

These updates mark a step forward in balancing creativity and responsibility. Sora 2 continues to evolve as a playground for artists, storytellers, and filmmakers, but now with guardrails that keep control where it belongs: in your hands.