Late last year, when the US government shutdown cut off SNAP benefits that low-income families rely on for groceries, videos on social media showed the result in frantic scenes. “Imma, keep it to yourself,” one black woman said in a viral TikTok post. Another black woman complained about taxpayers’ responsibility to her seven children with seven men, and another woman had a meltdown when her food stamps were rejected at the corn-dog counter.
Visible watermarks marked some videos as AI-generated – apparently, it was too difficult for racist commentators and hustlers to believe that the frenzy was real. “You’ve got people treating it like a side hustle, selling stamps, abusing the system,” complained conservative commentator Rich Odom. fox news Reported on Snap deepfake as if they were authentic before release corrections. Newsmax anchor Rob Schmidt claimed People were using the snap “to get their nails done, weaves and hair done”. (A basic fact is lost in the outrage: White Americans to complete 37% of SNAP’s 42 million beneficiaries.)
The fake videos are just pieces of a widening mosaic of digital blackface, a pattern that has increased over the past two years as generative AI video tools have become widely accessible. “There’s been a huge uptick in this,” says Safiya Umoja Noble, a UCLA gender studies professor and author of Algorithms of Oppression, which focuses specifically on digital biases against black women. “Digital blackface videos are actually drawing from the same racist and sexist stereotypes and tropes that have been used for centuries.” The net effect is a symbol of blackness stripped of cultural responsibility or stewardship – minstrelsy in short.
Coined in a 2006 academic paper, the term digital blackface describes a form of Black cultural commodification repurposed for non-Black expression online. Examples range: posts in African American vernacular English, use of dark-skinned emoji, reaction memes featuring Beyoncé, Katt Williams, and other examples of Black cool.
“Early research on digital blackface began with white gamers using Bitmoji of a different race and changing their vernacular to represent themselves,” says Mia Moody, a Baylor University journalism professor whose upcoming book, Blackface Memes, examines the role of black users in starting and spreading online trends. “It’s part of cultural appropriation, gaining cultural capital. You might be a bland white person, but if you use this fantastic embodiment of a dark-haired black person, people will respect you. You’re suddenly interesting.”
As memeology has expanded into short-form videos, Black expression has increasingly been separated from authorship, context, or consequence. Internet culture scholars say some non-white online creators use AI-generated avatars based on familiar black faces – beauty influencers, culture podcasters, street interviewers; They pop up in a feed with real Black content creators. Big language models scour digital spaces that have gained cachet from Black speech and humor, absorbing their tone and slang. Hume AI is one of several companies offering synthetic voices for podcasts and audiobooks such as “a black woman with a subtle Louisiana accent”, or “a middle-aged African American man with a tone of hard-earned wisdom”. In most cases, creators whose speech is removed from YouTube, podcasts, and social media receive no compensation, not even knowing that their personalities have shaped these models.
However, the snap reaction clip was a notable addition to the mainstreaming of digital blackface – less blending, more weapons-grade stereotyping. Many of those videos were created with OpenAI’s text-to-video app Sora. As Sora’s popularity grew in 2025, users took advantage of its surrealism to tarnish the image of Martin Luther King Jr., sparking ethical debate around “synthetic resurrection”. Deepfakes showed him stealing from a store, wrestling Malcolm X, and swearing through his I Have a Dream speech. Conservative influencers flooded feeds with an AI-generated embrace between King and Charlie Kirk, conflating their conflicted legacies and cultural martyrdom. Bernice King, MLK’s daughter and director of his Atlanta-based nonprofit organization, criticized the slogan as “silly.”
Inevitably the Trump White House has joined in on the action. In January the official White House An image depicting Obama as a monkey was circulated via Trump’s own Truth social account earlier this month.
Blackface remains at the bottom of American mass media, even as it continues to grow at an alarming rate. Its roots are found in minstrel revues of the early 19th century; White artists applied grease paint made from burnt cork to their faces and plastered oversized white lips to depict black faces, and performed exaggerated routines of black laziness, clownishness, and hypersexuality. Thomas D. Rice, a Manhattan playwright, became famous in the 1830s by playing a foolish trickster named Jim Crow – a name that soon became shorthand for the forced racial segregation policies in the American South that persisted until the Civil Rights Act of 1964.
In their heyday, minstrel shows were The dominant form of American entertainment – reflected in newspaper cartoons and the highly popular Amos ‘n’ Andy radio show. After the Civil War, in order to gain a foothold on the stage, black performers were once again forced to adopt minstrel elements, at the expense of their individuality. “The objectives were, first, to earn money to help educate their young children, and second, to try to break down the bad feeling that existed against colored people,” Explained Tom Fletcher, a vaudeville singer and actor for nearly 70 years, died in 1954.
Even as minstrelsy disappeared from the headlines in the early 20th century, its toxic remnants persist in American culture – from the roving crows of Disney’s Dumbo, to Ted Danson’s infamous 1993 blackface roast of Whoopi Goldberg, to the annual parade of white Halloween revelers in racial garb. A decade ago, when the Internet was still a kind of black box, researchers like Noble and MIT Joy Buolamwini There were warnings about implicit racial biases in the coding of algorithms related to medical treatment, loan applications, hiring decisions and facial recognition. Now it’s out in the open, a stain wider and darker than any burnt cork routine.
Tech companies have made some efforts to stem the tide of digital blackface. Bowing to the public backlash, the King family and more prominent estates, OpenAI, Google, and AI image generator MidJourney rejected deepfakes of King and other American icons. In January 2025, Meta removed two of its own AI blackface characters – Grandpa Brian and a retiree named Liv, described as a “proud Black queer mom” and a “truth teller” – following allegations from their non-diverse development team. storm of criticism. Instagram and TikTok and others have made some efforts to crack down on viral digital blackface videos, For lackluster results. Last summer, attempts to replicate the Bigfoot baddie — an AI incarnation of a black woman as a human-Yeti hybrid in a pink wig, acrylic nails and hair bonnet created by Google’s Veo AI — turned into a full-on frenzy on social media, with some users even starting how-to courses. Avatar is still on social media.
Black in AI and the Distributed AI Research Institute (DAR) are among a handful of affinity groups that have emphasized diversity and community input in AI model-building to address programming bias. The AI Now Institute and the Partnership on AI have highlighted the risks of AI systems learning from data from marginalized communities and said tech companies can provide mechanisms such as data opt-outs to help limit harmful or exploitative use. But widespread adoption has been slow.
“About 400 hours of content is uploaded to YouTube alone every minute,” says Noble. “With the AI generation, these tech companies can’t manage what’s coming through their systems. So they don’t do it. Or they do what is absolutely necessary for the US government. But if you have an authoritarian regime in power, they can use your systems to facilitate propaganda.”
Although it is difficult to measure the exact impact of AI-generated digital blackface, its use by the Trump administration highlights its potential as a powerful tool of official disinformation. The Obama Truth Social entry revived a stigma that has pervaded dark online corners for years, and one that coincides with Trump’s continued efforts to discredit the former first family. (Trump denied direct responsibility and declined to apologize for the post, which was removed.) Meanwhile, Armstrong’s doctored image at the White House, replaced by a real photo taken by the Department of Homeland Security and published on his official Twitter account, has been scanned by the government as a psychiatrist working closely with tech companies to monitor activists and other perceived enemies of the state.
In addition to denouncing bigotry as news, digital blackface exposes black users to a level of personal abuse and harassment that is reminiscent of an era when racists were completely empowered to express their bigotry without permission. And then, as now, it seems little can be done to curb the poisoning. “We are living in the United States with an open, no-holds-barred, anti-civil-rights, anti-immigrant, anti-Black, anti-LGBTQ, anti-poor policy agenda,” Noble says. “Finding the material to support this position is a matter of bending reality to suit the state’s own imperatives. And that’s easily done when every tech company stands behind the White House.”
Still, Moody expects the current fascination with digital blackface will soon become as outdated and distasteful as the analog variant. After all, he has seen this drama before. “Right now people are experimenting with AI technology and seeing what they can achieve,” she says. “Once we get past that, we’ll see less of it. They’ll move on to something else. Or they’ll be ready for the job, and that would be a shame. Just look at history.”
