Deepfake detection models utilizing subject-specific libraries

This invention describes a system that can tell if a piece of digital content, like a video or audio clip, is a deepfake. Instead of using one general detector, it keeps a separate, specialized deepfake detection model for each person or subject it's trying to protect. When someone asks if content related to a specific person is fake, the system uses that person's unique model to analyze the content and then alerts them if it looks suspicious.

Why it matters: Since 2021, the widespread availability and sophistication of generative AI tools have dramatically increased the ease and quality of deepfake creation. This heightened threat makes robust, potentially subject-specific, detection systems like this more critical and valuable now than when it was filed.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedMachine learning model training and deployment

3 AI reinvention directions for this patent

See all three

AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.

Reinvent this with AI

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10