AI-Generated Disinformation Becomes Routine, Undermining Public Trust
Data from European media monitoring organizations show that AI-generated synthetic media—deepfakes and manipulated content—have surged dramatically since September 2025, becoming routine in disinformation campaigns tied to major news events. The proliferation of accessible image and video generation tools enables increasingly credible false content that can alter public perception and undermine trust in authentic evidence.
What happened
AI-generated synthetic content has become increasingly integrated into disinformation campaigns between September 2025 and April 2026, with montages and deepfakes moving from exceptional to routine in news flows.
Data from the European Digital Media Observatory (EDMO) show sustained increases in AI-manipulated content used for disinformation in recent months. In March 2026, 34% of debunked content by EFE Verifica involved synthetic media, the highest recorded percentage.
Releases of more realistic and accessible image and video generation tools since September 2025—including Ray3, Sora 2, Veo 3.1, and Firefly Image Model 5—enable creation of false images, videos, and audio that can be adapted to current events.
Reported impact
AI-generated disinformation alters public perception of reality by fabricating increasingly credible content tied to current events such as the Iran conflict, Maduro's capture, and the Epstein case.
A secondary effect has emerged where videos or images are accused of being synthetic to discredit authentic evidence, undermining public trust in authentic media.
Index analysis
The timing of tool releases from September 2025 onwards coincides with a marked increase in synthetic content reaching peak proportions in March 2026, showing a direct correlation between tool availability and disinformation volume.
AI-generated content exploits information vacuums and real events to manipulate narratives. Content also includes 'AI slop'—sensationalist material designed for clicks and monetization rather than specific propaganda goals.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.