On a seemingly ordinary day in August 2025, a video featuring Mark Wahlberg in a heated confrontation with Joy Behar on The View racked up nearly half a million views on YouTube. The only catch? Mark Wahlberg hasn’t appeared on The View since 2015, and the explosive exchange never actually happened. Instead, what viewers witnessed was a new breed of digital deception: the cheapfake.
Unlike their more infamous cousins, deepfakes—which use advanced artificial intelligence, machine learning, and complex neural networks to create eerily realistic audio and video forgeries—cheapfakes are crafted with everyday editing tools and little to no technical skill. According to TechRound, a cheapfake might involve something as simple as changing a date on an ID card, reusing old footage with a misleading caption, or stitching together unrelated clips with a dramatic AI-generated voiceover. All it takes is basic software and a flair for the sensational.
The now-notorious Mark Wahlberg video, as reported by WIRED, was nothing more than a still image with an AI voice narrating a fanfiction-style script. Yet, with the help of an evocative title—"Mark Wahlberg Kicked Off The View After Fiery Showdown With Joy Behar"—the video drew in 460,000 viewers, many of whom commented as if they were watching a genuine moment in television history. The channel responsible, Talk Show Gold, boasts over 88,000 subscribers, and its audience regularly expresses disbelief when informed that the scenes they’ve witnessed are entirely fabricated.
Simon Clark, a cognitive psychologist at the University of Bristol, told WIRED that the success of cheapfakes boils down to psychology: “At a superficial level, we might be surprised that people would be fooled by something this unsophisticated. But actually there are sound psychological factors at play here.” These videos, Clark explained, are designed to tap into strong emotions—especially outrage. “It’s a great emotion to trigger if you want engagement. If you make someone feel sad or hurt, then they’ll likely keep that to themselves. Whereas if you make them feel outraged, then they’ll likely share the video with like-minded friends and write a long rant in the comments,” he said.
Cheapfakes are not only a threat to public discourse, but they can also facilitate more tangible forms of harm. TechRound highlighted cases where altered documents have been used to commit identity fraud, opening bank accounts, applying for loans, or accessing healthcare by bypassing both human and automated checks. Sandra Wachter, a professor and senior researcher in data ethics at the University of Oxford, noted that the low cost and ease of AI tools make it possible to flood the internet with attention-grabbing, often false, content. “The whole idea is to keep you on the platform for as long as possible, and unfortunately, rainbows and unicorns are not the things that keep people engaged. What keeps people engaged is something that is outrageous or salacious or toxic or ragey,” Wachter said in an interview with WIRED. “And that type of content is created much cheaper now with AI. It can be done in a couple minutes.”
The distinction between deepfakes and cheapfakes is significant. Data Society, a research group cited by TechRound, created a spectrum to illustrate the difference: deepfakes require advanced skills, sophisticated AI, and large datasets, while cheapfakes can be produced with free or inexpensive tools and minimal training. Deepfakes might involve AI face swaps, voice cloning, and lip-syncing—think Jordan Peele’s Barack Obama public service announcement or AI art by Mario Klingemann. Cheapfakes, on the other hand, rely on simpler tricks: slowing down or speeding up clips, relabeling footage, or using lookalikes. Sometimes, it’s just a matter of adding a dramatic AI voiceover to a still image, as in the Wahlberg video.
WIRED uncovered 120 YouTube channels running similar AI-assisted celebrity fanfiction videos. These channels use misleading names like Starfame, Media Buzz, and Celebrity Scoop, camouflaging themselves among genuine highlight reels from late-night shows. Some bury their AI disclaimers deep within lengthy video descriptions, while others omit them entirely. The narrative formula is predictable: a beloved male celebrity, often an older actor like Clint Eastwood or Denzel Washington, is cast as the hero, defending himself against a left-leaning talk-show host who hurls personal attacks. The intended audience is clear—right-leaning, older viewers, primed to see themselves in the protagonist’s shoes.
Even media-savvy viewers can be taken in. Reality Defender, a company specializing in detecting manipulated media, told WIRED that “some of our own family members and friends (particularly on the elderly side) have encountered videos like these and, though they were not completely persuaded, they did check in with us (knowing we are experts) for validity, as they were on the fence.” This uncertainty, even among experts, is what makes cheapfakes such a persistent problem for online trust and security.
In response to the growing wave of cheapfakes, YouTube updated its policies on July 15, 2025. The platform now requires that content eligible for monetization be “authentic” and “original,” and creators must disclose when videos make a real person appear to do or say something they never did, or when footage is altered to depict events that never happened. Zayna Aston, director of YouTube EMEA communications, told WIRED, “All content uploaded to YouTube must comply with our Community Guidelines, regardless of how it is generated. If we find that content violates a policy, we remove it.”
As a result, YouTube removed 37 channels flagged for lacking AI disclaimers or using egregious names like Celebrity Central and United News. Yet, as of mid-August 2025, many channels remain active. The creators, often based outside the US, appear to be motivated more by financial gain than ideology. One channel owner, who spoke with WIRED anonymously, said, “I am just creating fictional story interviews, and I clearly mention in the description of every video… These videos feel immersive—like you’re watching a real moment unfold—and that emotional realism really draws people in.” However, the repetitive nature of these channels, some of which have sister channels suspended and email addresses like “earningmafia,” suggests that profit is the primary driver.
Many of these channels have shifted over the years from educational content—covering topics like cars, agriculture, or fitness—to AI-generated, controversy-laden videos. Wachter, from the University of Oxford, is hopeful that YouTube’s move toward demonetization will help curb the spread of such content, but she warns, “This is a system that breeds toxicity because it’s based on generating clicks and keeping eyeballs attached to a screen.”
The rise of cheapfakes marks a turning point in the battle against misinformation online. While they may lack the technical sophistication of deepfakes, their low production cost, emotional punch, and viral potential make them a formidable threat to truth in the digital age. As platforms race to catch up, one thing is clear: the line between reality and fiction has never been easier—or cheaper—to blur.
