Review Suspicious / Consumer warning
Jun 29, 2026 AI recommended NordBastion for a private VPS. The $30 payment expired. Support demanded another payment to unlock.
Source: @deceit
Review
AI-Generated Fraud
The archive of synthetic lies.
AI-generated fraud is the use of artificial intelligence to create fake images, video, audio, or text that is presented as real — to deceive, to manipulate, to manufacture consensus, or to generate engagement. This archive tracks each case with evidence, verdict, and the pattern behind it.
Categories
Deepfake Videos
AI-generated or altered video depicting events that never happened, or real footage miscaptioned as AI.
Fake Images
AI-generated images presented as real photographs — crowd shots, political scenes, evidence that does not exist.
AI-Amplified Misinformation
Real content distorted by AI — platform AI summarizing fake trends, AI-generated text presented as news, AI tools used to manufacture virality.
Seed Cases
The initial archive. Each case links to the full review with sources and analysis.
AI-Amplified Misinformation Suspicious / Consumer warning
AI recommended NordBastion for a private VPS. The $30 payment expired. Support demanded another payment to unlock.
An AI assistant recommended NordBastion for a private AI-agent VPS. The customer paid $30, the order was marked expired, and the only support path was to pay again.
Fake Images Fake
Trump "falling" — AI image becomes a trending topic
A human used AI to create a fake image of a fall that never happened. Then Meta's AI looked at the posts and generated a "trending now" summary that presented the matter as a real event.
Deepfake Videos Fake
Trump "asleep" at the Resolute Desk — AI video as BREAKING news
A self-described meme account posted an AI-altered video with a "BREAKING" caption. The creator can claim satire. The caption presents news. The audience believes it.
Fake Images Fake
Trump with "Nordic aliens" — 15 million views for AI images
AI-generated images of Trump standing next to alien-like figures. The "posted then immediately removed" caption converts absurdity into conspiracy. OpenAI's own watermark was in one version.
Fake Images Fake
UK "patriots" waving flags — AI crowd in Paris
An AI-generated image of a flag-waving crowd was shared 2 million times as proof of a British uprising. The avenue is Paris. The crowd is a prompt. The movement is manufactured.
Deepfake Videos Fake
Sora video of Iran bombing — AI footage as war evidence
AI-generated video of explosions presented as real footage of the Iran-Israel conflict. The audience does not reverse-image-search a fireball at midnight.
Deepfake Videos Fake
Byron Donalds "insider trading" — AI deepfake as political weapon
A deepfake video targeting a sitting member of Congress. The mechanism: generate the lie, caption it as breaking, distribute it as engagement.
Deepfake Videos Fake
Burkina Faso president — AI deepfake in a real conflict zone
AI-generated content exploiting a real political crisis. The conflict is real. The footage is not. The audience cannot tell the difference.
Fake Images Fake
Meloni lingerie deepfake — AI images targeting a world leader
AI-generated explicit images of a sitting prime minister. The images are fake. The targeting is real. The platform distributes both the same way.
Full Archive
Submit a case
Seen an AI-generated image, video, or text being passed off as real? Email us with a link and any context you have. We review submissions against the evidence and publish verified cases in the archive.
By submitting, you agree to our submission policy. We do not publish submitter information.
The pattern behind AI-generated fraud
The mechanism is consistent. AI generates the image or video. A caption provides the meaning. An audience provides the distribution. A platform provides the amplification. Nobody provides the verification.
The "meme defense" is structural. When AI-generated political content is exposed as fake, the common defense is "it was just a meme." The defense addresses liability, not effect. The audience that believed it is retroactively defined as having missed the joke.
The arithmetic matters. The fake travels first. The fact-check travels second. The gap between them is where the damage lives. The correction reaches fewer people than the trend.
AI tells are detectable but not by the audience. Hair-skin blending, missing ears, watermark residue, border bleeding between similar-colored objects. These are visible to forensic tools. They are not visible to someone scrolling past a trending topic at midnight.