Cyabra is used by government agencies and Fortune 500 companies around the world to monitor billions of online conversations in real-time while detecting misinformation and disinformation campaigns. Leveraging machine learning to identify behavioral patterns of fake and real authors, the platform maps the spread of disinformation across online sources such as traditional media, social media, blogs, forums and open instant messaging groups.
Cyabra’s platform provides deep insights into the layers of these online conversations, helping organizations understand the rapidly changing conversations and their changing sentiments taking place across the web and on social media. Cyabra’s proprietary deep learning algorithms are the only proven technologies able to consistently scan, detect, analyze and GeoTag Generative Adversarial Networks (GANs), fake profiles, deepfake videos and deepfake still images. Cyabra’s authenticity rating algorithm analyzes over 600 different parameters to determine whether an online entity is real or fake with an average accuracy rating of over 95% and 85% accuracy in identify deepfakes.
Cyabra’s platform is a SaaS based model that can also be integrated via API (hosted on AWS and Azure). The system’s notification feature sends customizable alerts based on activity surrounding subjects of interest. In addition to government agencies, Cyabra is relied upon by enterprise marketing executives, financial analyst teams and newsroom media professionals to provide unique social insights into online conversations as they happen.
The graphic above is a breakdown of fake/real/bad authors, identified by Cyabra’s proprietary algorithms, with illustrations of authenticity reasons and activity levels.