How the internet destroyed our ability to tell truth from fake

Tools that made it possible to distinguish real photos from staged ones, document war crimes, and verify videos spread on social media have come under threat. Online content verification is experiencing a systemic crisis — and there are several reasons for it. Artificial intelligence has radically transformed the disinformation landscape over the past two years.

Generative models have learned to create images and videos of such quality that even media literacy experts cannot confidently distinguish them from real ones. AI content detectors exist, but work unreliably: they regularly make mistakes in both directions, falsely accusing authentic images and missing generated ones. Parallel to this, another problem is growing — the closure of satellite data.

Commercial providers that once openly provided images for independent investigations — from analyzing concentration camps to tracking military equipment — increasingly restrict access under pressure from governments or for commercial reasons. Journalists and NGOs that used satellite photos as a key documentation tool are losing a vital source. The problem is compounded by the speed of content spread.

By the time verification is complete, millions have already seen the fake. Social media algorithms are optimized for engagement, not veracity — and emotionally charged but false content spreads incomparably faster than corrections. Research shows that false news spreads six times faster than true news.

As a result, the average user gets the feeling that nothing can be trusted — neither photos, nor videos, nor texts. This itself becomes a tool of manipulation: when people lose trust in everything, they either sink into information nihilism or only believe sources that confirm their already established beliefs. What this means in practice: the race between tools for creating fakes and tools for detecting them will continue, but the distance between them is growing.

A solution requires not only technology, but also investment in media literacy, pressure on platforms, and openness from satellite providers. Until that happens — verification of reality remains a privilege of specialists, not a norm.