As the Flemish knowledge center for digital and media literacy, Mediawijs sees it as a core mission to strengthen citizens against mis- and disinformation. The VISAVIS project – fully titled VISual and AI Disinformation Verified with Information Literacy and Citizen Science – gave us a valuable opportunity to reinforce that ambition. At a time when image manipulation and AI-generated content are becoming increasingly convincing, it is more important than ever not only that people have access to reliable information, but also that they know how to critically assess images and messages.
VISAVIS started from a clear observation: visual disinformation undermines trust in our society and fuels polarization. This dynamic became painfully clear during the COVID-19 pandemic and in recent geopolitical conflicts, where for example (war) images were taken out of context or AI-generated visuals circulated as propaganda. In this project, we joined forces with partners from Belgium, Sweden, and Greece: IMEC-DICE (project coordination), Mediawijs (media and information literacy), Folkets Bio Filmpedagogerna (film and media education), and CERTH (technology partner).
Three pillars: technology, education, and citizen participation
The project rested on three strong foundations that reinforce each other:
- An online platform
At the heart of VISAVIS is a user-friendly online platform with two free tools to check visual content:
- The Image Checker, which analyzes whether an image may have been AI-generated or manipulated.
- The Video Checker, which helps determine whether a video has been taken out of its original context.
- A citizen science campaign
During an intensive two-month campaign, Belgian participants got hands-on experience. They collected suspicious images, analyzed them, and at the same time worked on their own skills through training and weekly exercises. - Educational material
We developed tailored training materials for participants, which we are now refining based on their feedback. Even after the project, these materials remain publicly available.
Participants as active researchers
Within the citizen science approach, participants took on an active role as “researchers.” They learned to analyze images, ask critical questions, and share their insights during the sessions. Soon it became clear that visual disinformation is not limited to AI-generated images—classic editing tools like Photoshop can also mislead. This insight underlined that critical viewing requires more than just technology. It is also about understanding context, intent, and visual language.
Strong engagement from older participants
What pleasantly surprised us: the campaign especially appealed to older participants (55+). This group was far from absent in the digital world. Many showed strong motivation to keep up and to further strengthen their skills—particularly around topics like online safety and reliability. Of course, some felt less digitally skilled, but they too seized the learning opportunities with both hands.
Interestingly, a significant portion of the group was in fact highly digitally skilled. For some, the trainings were even a bit too basic, which highlighted an important insight: the stereotype of older people “not keeping up” is far from accurate. This generation is diverse, with varying digital experiences and skills, and precisely that offers opportunities to continue involving them in digital resilience in a meaningful way.
Reaching young people (18–34) proved more difficult. Yet they are active on the platforms where visual disinformation circulates most. Still, we made progress: thanks to a collaboration with Artevelde University of Applied Sciences, we engaged a group of students through the “factchecking” course. A fine example of how young people can indeed be motivated when the context aligns with their daily lives.
Practical trainings and weekly exercises
Participants in the Belgian campaign followed three hands-on training sessions.
- In the first session, they were introduced to the platform and learned to work with the Image and Video Checker.
- The second session went deeper into how we consume information, and the role of images and AI in that process.
- The third session focused on the mechanisms and motives behind disinformation.
Each session was offered three times—once in person and twice online—so that everyone could join at a convenient moment. Sessions lasted about 90 minutes and were guided by Louis and Zara from Mediawijs, both experienced in news and information literacy.
In addition, participants received a weekly exercise via the Qualifio platform. Assignments varied: from recognizing manipulated images, to finding the context of a video, or assessing the motives behind a disinformation campaign. Some tasks took just fifteen minutes, others a bit longer. After each exercise, participants received a clear solution with extra explanation. This way, they gradually developed their skills and gained a broader perspective on visual disinformation.
Looking critically at the tools
The platform tools sparked not only enthusiasm, but also valuable questions. The Image Checker proved popular: its probability score (from “weak indications” to “very strong indications”) offered guidance and encouraged reflection. At the same time, participants wondered how exactly that score was calculated and what factors played a role.
The Video Checker works differently: it does not provide a score but shows the results of, among other things, a reverse video search. Here, the user must draw their own conclusions. This stimulates critical thinking, but for some participants requires extra practice.
Working together toward a resilient society
After the Belgian campaign, many participants reported feeling more confident in recognizing visual disinformation. This increased media literacy can, in the long term, contribute to a society that engages with information more critically and consciously. And that is exactly what VISAVIS and Mediawijs aim for: empowering citizens so they can actively, critically, and consciously navigate the digital world.
The relevance of the project is crystal clear. In a time when AI content is hardly distinguishable from reality, a new challenge arises. What is the greatest risk: that we can no longer tell fake from real, or that we start distrusting everything—even what is true? VISAVIS advocates vigilance, but not cynicism. Critical thinking, without falling into distrust.
Lasting impact and the road ahead
The Belgian campaign left lasting traces:
- The platform is being improved based on participants’ feedback.
- The trainings are reusable in new educational contexts.
- The knowledge built up is being shared, also beyond Belgium. There are plans to introduce the VISAVIS platform in other European countries together with our international partners.
In short: VISAVIS showed how technology, education, and citizen participation can reinforce each other. By jointly investing in media literacy, critical thinking, and active participation, this project made a valuable contribution to a more resilient and informed society. For Mediawijs, VISAVIS seamlessly fits our mission: empowering people in an ever-changing digital world.