Creating Disinformation

Techniques 

• Manipulating existing images or videos, creating “deep fakes,” synthesizing text, and using memes or parody that may be misinterpreted as real.

Technologies 

• Range from image editing tools to sophisticated machine learning technologies such as GANS (Generative Adversarial Network) or language models. 

Spreading disinformation

Techniques 

• Targeting specific individuals or populations by characteristics, false social media accounts (fake personas, bots, cyborgs, hacked accounts), networks within and across platforms, tricking mainstream media and journalists.

Technologies 

• Data acquisition and aggregation, bots and botnets, network building, shifting content from positive or centrist to extremist views through NLP, language models and synthesized text.
   

Detecting and Deterring Disinformation

Research addresses built-in affordances and economic incentives of social media platforms such as recommendation and trending systems, cross platform connectivity, ad placement, account anonymity, and the speed of disinformation spread. Researchers also investigate the potential for machine learning to detect fakes, and methods to deprecate the visibility of disinformation content.