Microsoft launches deepfake detector ahead of US election

Ahead of the November 2020 U.S. election, tech giant Microsoft has launched a new technology that could spot and expose synthetic (manipulated) media, otherwise called deepfakes.

The new tool, according to Microsoft, could analyse videos and still photos to generate a manipulation score, a development that would be useful in detecting fraudulent videos and photos in the event of use by any political player.

As deepfakes have garnered widespread attention for their uses in celebrity pornographic videos, political exhibitions, revenge imaging, fake news, hoaxes, and financial fraud; the new tool is expected to elicit responses from both industry and government.

READ ALSO: Microsoft obtains court order to seize web domains used by South Korean hackers

Commenting on the design of the tool, Microsoft stated that the technology is capable of providing its monitoring team “a percentage chance, or confidence score” that the media has been artificially manipulated.

Microsoft stated: “In the case of a video, it can provide this percentage in real-time on each frame as the video plays.

“It works by detecting the blending boundary of the deepfake and subtle fading or greyscale elements that might not be detectable by the human eye,” the company added.

Join the conversation

Opinions

Support Ripples Nigeria, hold up solutions journalism

Balanced, fearless journalism driven by data comes at huge financial costs.

As a media platform, we hold leadership accountable and will not trade the right to press freedom and free speech for a piece of cake.

If you like what we do, and are ready to uphold solutions journalism, kindly donate to the Ripples Nigeria cause.

Your support would help to ensure that citizens and institutions continue to have free access to credible and reliable information for societal development.

Donate Now