Russian Disinformation Video Used AI to Fabricate a USAID Celebrity Payoff Scheme
What happened
In May 2025, an AI-generated video began circulating online with a specific and false claim: that the U.S. Agency for International Development had paid Hollywood celebrities millions of dollars to travel to Ukraine and shore up public support for President Volodymyr Zelensky among American audiences. The video was fabricated. No such payments happened. Fact-checkers and media outlets identified the claims as false within days of the video's spread.
The video's design borrowed from a familiar disinformation template: give the claim a face, attach a dollar figure, and let outrage carry it further than any rebuttal can travel. By attributing fictional payments to a U.S. government agency and connecting them to a politically divisive figure, the video placed itself directly inside ongoing American debates about foreign aid and the Ukraine war. The AI-generated format made celebrity voices and appearances easier to fake convincingly enough that casual viewers would not immediately recognize them as fabricated.
The video was not a standalone piece of content. It was one output of a coordinated Russian government-linked disinformation campaign known as Operation Overlord, also tracked under the designations Storm-1679 and Matryoshka. That campaign has used AI tools to produce content designed to look authentic and circulate in Western media environments, mixing genuine controversy with fabricated evidence to make the fabrications harder to isolate and dismiss.
The damage extended beyond the video itself. The disinformation contributed to a measurable surge in conspiracy theories and hostile sentiment directed at USAID, arriving precisely as political forces in the United States were already pressing to dismantle or significantly reduce the agency's operations. The false claim did not need to be believed by everyone to be effective. It needed only to add noise, deepen suspicion, and give existing critics a shareable piece of content that appeared to confirm what they already suspected.
What this incident makes visible is a verification gap that AI-generated disinformation is designed to exploit. A fabricated video spreads through the same channels as legitimate reporting, carrying the same surface characteristics: a voice, a face, a number, a claim. Without a provable record of what a system produced, when it was created, and what its actual sources were, the correction always arrives after the damage and competes for space with the original lie. The infrastructure for tracing AI-generated content to its origin and confirming what is real has not kept pace with the tools available to manufacture what is not.
Reported impact
- Affected parties
- Not publicly disclosed
- Harm type
- Not publicly disclosed
- Scale
- Not publicly disclosed
- Financial impact
- Not publicly disclosed
- Regulatory action
- Not publicly disclosed
Classification
Relevant governance controls
Governance control mapping is not available for this record.
- No controls mapped
Not publicly disclosed
Control mapping is analytical. It does not state that any control would have prevented the incident.
Sources and evidence
This record was researched and written by the Index. The event is also catalogued in the following database, which is listed for cross-reference.