For a decade, defending democracies against influence operations has meant reacting after the damage was done. Since Russian troll farms first weaponised social media around the 2016 US election, the standard response to influence operations has been reactive: detect, debunk, take down, repeat. By the time a campaign was identified, it had usually already done its work.

That dynamic is finally changing.

In a recent interview with Danish outlet A4 Medier, LetsData co-founder Ksenia Iliuk argued that AI, often framed as the great accelerator of disinformation, also gives defenders something they have never had before: the ability to see coordinated campaigns forming before they spread.

"This is the first time we have a chance to be one step ahead. Before, the intelligence services have had to play whack-a-mole to stop influence operations," Ksenia says.

The stakes are not theoretical. Russia is estimated to have spent €400 million on influence operations around Moldova's parliamentary elections last year. States do not invest at that scale without expecting returns.

Read the full interview in A4 Medier.