Pundit by Prompt
The Mass Production of AI Personas Weighing In on World Affairs
Overview
We discovered over 300 assets across YouTube, Facebook, TikTok, X, and more niche online spaces, posting and promoting video and audio content featuring recurring AI personas or AI deepfakes of public figures. These accounts are posing as experts on a range of economic, social, and political topics, primarily in English and Spanish. In addition, they are attracting views and subscribers – who sometimes even engage with comments and questions, even though most of the YouTube videos we found are labeled AI-generated, -altered, or -dubbed. We didn’t observe such labels on this content on other social media platforms.
We set out to research these assets and their content, to understand the actors, behaviors, and intent behind them, and analyze potential signs of influence operations and inauthentic coordination. Although we couldn’t identify the operators of these assets, or confirm that they’re linked, we did determine that a shared content-generation infrastructure and amplification strategy is likely: Many of the accounts and channels post content featuring the same personas discussing similar topics. We also determined that monetization may be a motive behind some of these assets.
Key Findings
Recurring AI Personas
Between May and June 2026, we identified over 150 assets on YouTube, Facebook, TikTok, X, and other platforms, using a recurring set of AI personas to disseminate geopolitics and world affairs commentary, mainly in videos. The personas also discuss mainstream news, technology, the global economy, finance, investment, and gold and silver, among other topics, in English, French, Korean, Portuguese, and Spanish.
Deepfake Impersonations
In addition, more than 140 YouTube channels are hosting videos featuring AI deepfakes that impersonate prominent journalists, academics, and business leaders, such as Col. Douglas Macgregor, Jiang Xueqin, and Rachel Maddow, commenting on U.S. politics and global events.
Anti-West Bias
At the time of publication, over 200 of the identified YouTube channels had been removed. Based on open-source information, removal reasons included violations of Google’s or YouTube’s Terms of Service or YouTube’s community guidelines.
Channel Timeline
Although much of the discussion in both types of video is factual and unbiased, anti-West and pro-China commentary is a recurring theme. The basis is often economic and political events involving the U.S., including the conflict in Iran, the Russia-Ukraine war, and ongoing tensions in the South China Sea region. Most channels were created in 2026, but others date back to 2007. Several channels we documented during our research have since become unavailable, and others have deleted old content, replacing it with new videos and shorts that may feature a new topic of focus or a different AI persona.
Uncertain Coordination
We could not confirm that the identified assets are part of a coordinated effort and/or the same network. However, they exhibit the same behavior, amplify the same videos, and use the same AI-persona and video-template styles. We could not identify the origins or the likenesses of the AI personas through open-source research.
Monetization Signals
So prominent is one of the AI personas, featured in 154 channels, that social media users have taken notice of its appearance in spammy, low-effort videos discussing silver, gold, finances, and investment, and have even created a meme based on it. The prevalence of this character and others indicates that some of the assets could belong to a network operated for monetization, intended to gain maximum views – and succeeding.
Disclosure Labeling
Most of the YouTube videos we observed have a label indicating that content has been meaningfully altered or synthetically generated; creators are required to disclose such content if it features realistic synthetic people, voices, places, or events. We didn’t observe this type of label in the content we found on other platforms.
Dina Sadek delivers customized monitoring, in-depth analysis, and investigations to diverse Graphika clients. Her research expertise includes open-source intelligence, analysis of online information environments and narratives, and identifying influence operations and inauthentic coordination on social media platforms, among other online harms.
Lili Turner specializes in tracking and analyzing Chinese influence operations targeting global audiences and democratic processes. Additionally, she has expertise in tracking AI-enabled inauthentic behavior, platform abuse, and nation-state cyber threats. At Graphika, she has led investigations into networks of fake news domains and threats to commercial interests.
Download the complete PDF
The full report includes the complete network graph maps, raw attribution indicators, cross-platform topology analysis, and the full takedown timeline with platform-level data.
- Full network graph visualizations
- Attribution indicators with confidence scores
- Raw behavioral modeling data
- Takedown coordination timeline
Related Reports
Ore Else
Spamouflage Targets Chile’s Copper Smuggling Coverage
Read the Report
The #Americans
Chinese State-Linked Influence Operation Spamouflage Masquerades as U.S. Voters to Push Divisive Online Narratives Ahead of 2024 Election
Read the Report
Cheap Tricks
How AI Slop Is Powering Influence Campaigns
Read the ReportSee How Graphika Can Help Your Team Act on This Intelligence
This report is one of 600+ investigations Graphika’s team has published. Our platform gives your analysts continuous access to the same intelligence — plus the tools to apply it to your specific threat environment.