Teen Wins National Award For Inventing Deepfake Video Detector

Deepfake videos are a concern for many, primarily because they use AI to distort reality and can damage people in the process.

An Irish teen has now found an innovative solution to the problem. Greg Tarr, from County Cork, won the 2021 BT Young Scientist & Technologist of the Year award for his work to tackle deepfake videos, a project he calls ‘Towards Deepfake Detection,’ and he has plans to continue developing his technology.

Deepfake videos have been used to create pornography without consent by essentially swapping the face of adult performers with another person. There have also been misinformation campaigns, one of which shows President-elect Joe Biden falling to sleep in an interview, which of course never happened. Due to these uses, many want to regulate the AI technology more thoroughly.

Seventeen-year-old Greg Tarr explained his deepfake finding technology to Euronews:

I’ve been working on AI for maybe four years and it’s being trained to look at vast amounts of data, it’s a concept that is currently being done, but mine is ten times faster.

It’s amazing how few people understand how prevalent deepfakes are in current media.

The technology uses AI to detect the AI that has been used to create a deepfake. The impressive, and slightly ironic, methodology already delivers quicker detection than its competitors. Although Tarr did not go into too much detail about what process he used to achieve quicker detection, the teenager did note that he would like to push it further.

Speaking about his plans for the technology, Tarr said:

I look forward to representing my country in September with the same project or maybe an improved version, and I’m also looking forward to commercialising or ‘productising’ this development.

In terms of turning the project into a product, it seems that plenty of people would be interested in the technology that could stop the spread of misinformation. At the moment, authorities are attempting to combat the issue with federal laws, but an AI detection technology could mitigate the issue.

For now, it seems that Greg Tarr will return to working on his award-winning project, and it will be interesting to see how it develops. With continued hard work, Tarr could be at the forefront of deepfake detection in the future.

If you have a story you want to tell, send it to UNILAD via [email protected]