“ImageAware": Using artificial intelligence to detect disinformation online

Social media has given rise to a new form of hybrid warfare that is being used by conflict parties around the world: the dissemination of fake or manipulated content. How can artificial intelligence (AI) help to identify disinformation online and evaluate it for situational awareness? The Bundeswehr Cyber Innovation Hub (CIHBw) has developed a solution for this with its “ImageAware” innovation project.

The Russian war of aggression in Ukraine, the UN mission in Mali or the Hamas terrorist attack on civilians in Israel - today's conflicts are also being fought in the cyber and information space. One aspect of hybrid warfare is the targeted dissemination of disinformation in order to influence public opinion. Images, memes and videos in particular can generate negative emotions and are therefore used in a targeted manner to create a public that opposes our soldiers in action and German security and foreign policy. The authenticity of the content is difficult to verify.

The CIHBw has taken up this challenge. With the “ImageAware” innovation project, the CIHBw has developed a program together with the startup “Munich Innovation Labs” that uses AI to automatically capture and evaluate photos, videos and memes. The software is able to analyze large amounts of data in a structured way, helping analysts to find manipulated or fake posts and identify disinformation campaigns at an early stage.

“Images as part of war propaganda and disinformation campaigns are not a new phenomenon,” says Sven Weizenegger, Head of the CIHBw. “What is new, however, is the quality and speed with which manipulated or falsified images can be disseminated using digital technology. For us, this means we have to be faster. And we can do this through the clever use of digital innovations.”

fullscreen
© CIHBw

Evaluation of memes has been very time-consuming so far

Through the targeted placement of emotionalized media content by actors in the information space, the algorithms of social media platforms are used to achieve a wide reach - regardless of the authenticity of the content.

This is because the origin, veracity and authenticity of video, text or audio files are difficult to trace, especially in conflict situations. However, manipulated or falsified content can have serious consequences for trust in democracy and the tasks of the Bundeswehr: They raise doubts about the legitimacy of possible deployments and can weaken support for the Bundeswehr. This poses an immediate danger for soldiers in conflict zones.

To counter this threat at an early stage, the Bundeswehr also analyzes publicly available data from the internet. However, the evaluation of image and video material is particularly time-consuming. Until now, there was no AI-supported process that could capture and evaluate these large volumes of data in a structured manner.

With “ImageAware”, the CIBHw provides software that can solve this problem. The innovation project minimizes the time required, improves the quality of the analyses and contributes to a better picture of the situation. As a result, the Bundeswehr analysts have more resources for assessing the content. They can create a more comprehensive and detailed picture of the situation. This gives them an information advantage in a conflict environment and reduces the risks for soldiers in the area of operations.

LinkedIn Instagram Youtube