How do you prevent malicious/negative use?

We take the potential issue malicious/negative use extremely seriously. We have very strict content guidelines that users of our platform have to agree to before creating any kind of video with Synthesia.

Furthermore, here are a few ways we avoid any theoretical possibility of malicious/negative use on our platform:

  • 👔 Business videos only: our platform is for business use only. Our systems and content moderation teams check video scripts and remove any non-business content.
  • 📜 Script moderation: we have automatic and manual procedures to detect any kind of political,  sexual, personal, criminal or discriminatory content.
  • 🧒 Custom avatar moderation: each custom avatar goes through our approval flow, which means it's impossible to create an avatar that isn't connected to user's business. We require specific and expressed consent from each person that is presented as a custom avatar.

We are also actively working with industry partners to ensure this technology is used for responsible business purposes only. For example, Synthesia co-founder Matthias Niessner recently published FaceForensics++ a state-of-the-art system for deepfake detection, developed in his lab at TUM.

Read more on this topic from Synthesia: