Aos Fatos uncovers a digital deception that combines artificial intelligence with colonial mythmaking, all in the name of profit
What happens when ancient wisdom is faked by machine learning and monetised on social media? In this in-depth investigation, Aos Fatos exposes how false Indigenous personas that are generated by artificial intelligence and promoted through Facebook ads are being used to sell unverified “natural cures” to a largely Brazilian audience, manipulating both cultural imagery and public health discourse.
The piece begins by presenting the profile of a supposed cacique, or leader, of the Tupinambá – a man with facial features generated by AI tools and given a backstory as a wise Indigenous elder offering herbal remedies. This particular individual, as represented, does not exist; but the ailments his supposed cures promise to fix are very real for the vulnerable communities being targeted. The scam doesn’t end with the generation of a fake persona. The products sold in his name are unregulated, unproven and potentially dangerous.
According to Aos Fatos, these ads rely on longstanding colonial tropes: the “noble savage”, the mystical healer, the proximity to nature. They are all reappropriated and distorted by marketing teams who use AI not only to generate faces but also to write testimonials, create fake user reviews and even simulate customer service chats. “It’s a complete fictional ecosystem built with one goal: profit”, the authors note.
The ads, which appear to be professionally produced and often cite fake scientific studies or invented university endorsements, have reportedly reached tens of thousands of users. Some use manipulated or stolen images of real people; others invent characters entirely. All rely on Facebook’s ad infrastructure to hyper-target audiences by age, health interests and region.
A key revelation of the investigation is how these scams exploit the limited digital literacy of many users. “We spoke to elderly people who believed they were receiving advice from a real healer,” says one researcher quoted in the article. “They ordered these products thinking they were helping their arthritis or diabetes, but there was no medical oversight, no real guarantee of safety.” In some cases, users were charged repeated hidden fees or locked into subscription traps with no clear opt-out.
The article also highlights the complicity or indifference of platforms like Facebook, where these ads continue to circulate despite repeated user reports and the platform’s supposed content moderation protocols. In a damning critique, Aos Fatos outlines how Meta’s algorithmic moderation has repeatedly failed to catch these AI-generated deceptions, even after independent watchdogs flagged them.
AI ethics specialists interviewed in the piece warn that this type of content is only the beginning. “We’re entering an era where identities, especially those from marginalised or colonised cultures, can be entirely fabricated and commercialised by those with the right tools,” one expert notes. The threat is not only financial or medical, but cultural: when AI is used to simulate Indigenous speech, dress and beliefs, it erases real communities and replaces them with digital caricatures.
Some of the language used in these campaigns deliberately mimics that of traditional medicine, blending “ancestral knowledge” with pseudoscientific claims. One product claims to “rebalance energy fields” and “reverse chronic fatigue” through an herb supposedly harvested only in sacred forests – forests which, the investigation notes, do not actually exist.
Beyond the deception, the investigation brings attention to the broader public health implications. By positioning these fake cures as alternatives to medically-proven treatments, the ads risk steering people away from professional care. Brazil’s health system is already strained, especially in remote or underserved regions, and the proliferation of these AI-boosted scams only deepens mistrust between communities and health providers.
The article ends with a call for stronger digital regulation, platform accountability and media literacy education. “As AI technologies become more accessible, so too will the ability to fabricate trust. If platforms won’t police this, civil society must,” the authors argue. They also advocate for more Indigenous voices in tech governance, noting that the communities being imitated had no say in how their images or myths were being used.
Ultimately, the story is about more than just fake ads: it is a stark portrait of how AI is being used to exploit cultural trust, health fears and platform loopholes for profit, with consequences that ripple far beyond the digital screen.
Original article (“Anúncios no Facebook criam falsos indígenas com IA para vender ‘curas naturais’ sem eficácia”) by Gisele Lobato was first published in Portuguese on 15 May 2025.
It is available here.
Aos Fatos is an independent Brazilian fact-checking and investigative journalism organisation founded in 2015. It specialises in exposing misinformation, holding public officials and tech platforms accountable, and fostering transparency through rigorous, accessible reporting.
Summary by ZN
