Russia doesn’t need propaganda. It just needs a deepfake…

Earlier this year, religious leaders across the United States began reporting that artificial intelligence was being used to impersonate them. Among those targeted was Catholic priest and podcaster Mike Schmitz, who reaches over a million followers on YouTube. In these incidents, fake versions of the priest appeared online, delivering messages he had never recorded and directing followers toward fraudulent solicitations.

What happened

The impersonations utilized artificial intelligence to create videos that looked and sounded like the religious leaders. By mimicking the appearance and voice of trusted figures, the perpetrators exploited the relationship these individuals had built with their congregations. While many of the initial impersonations identified appear to be designed for financial gain, the technology poses broader risks regarding foreign disinformation. Adversaries can leverage the high level of trust associated with religious figures to amplify existing domestic anxieties or conspiracy theories. By using synthetic media, actors can create sermons that were never preached or generate forged communications, making fringe interpretations appear as though they represent a widespread consensus. This method of influence seeks to bypass the skepticism typically applied to government or anonymous political messaging by framing content within a familiar, moral, and faith-based context. The tactic mirrors historical influence operations, such as the 2016 Russian "Army of Jesus" campaign, but utilizes modern generative AI to achieve a scale and level of personalization that was previously impossible. Experts note that such campaigns do not necessarily require the invention of new grievances; instead, they amplify existing fears to make movements appear larger and more credible, potentially leading to real-world consequences such as threats against workers or the disruption of local infrastructure projects.