The current face detection system in Seedance blocks not only real faces but also fully AI-generated or stylized characters, making the tool unusable for professional filmmaking workflows that require consistent characters.
I’m a filmmaker currently using Dreamina (Advanced), and I’m running into a major limitation that’s making the tool almost unusable for professional work. The face detection system doesn’t just block real faces (which I understand from a safety perspective) – it also blocks fully AI-generated or stylized characters. Even harmless prompts get flagged. Example: I generated a video → took the end frame → reused it as input → still blocked. So this isn’t about “real people” anymore. The system is clearly overblocking. The result: – no consistent characters – no narrative workflows – no filmmaking use I’m in touch with other creators facing the same issue, and some are already moving to other tools because of this. Is anyone else experiencing this? And more importantly – has anyone found a workaround? Because right now, this feels like a complete blocker for actual production use.