Deep Fake Detection Algorithm
panuwat @ stock.adobe.com
Supported by artificial intelligence, a deep-fake detection algorithm is trained to identify deceiving anomalies in digital media that the human eye cannot notice easily or organically. This software is continuously run in the background, automatically scanning generative text, video, images and voice audio accessed by the user. When a deep fake is detected, the software notifies the user, similarly to an antivirus alert from malicious activity. Trained to analyze natural human body movements, facial alterations, the synthesizing of fake audio, and inconsistencies in pixels, this AI-based tool spots inauthentic information that is nearly indistinguishable from the original content. This technology helps both citizens and institutions to look for original and truthful sources and combat image appropriation.
Gender Equality
Challenges
Detection tools may pique the interest of privileged groups of society, such as politicians or business owners, thus leading various communities prone to deep fakes.
Opportunities
Helps ensure that women targeted online are being "erased" from pornographic content.
