With AI video generation tools becoming more advanced and widely available, so too are deepfakes. What once required vast technical know-how and/or expensive software can now be created by almost anyone with access to a consumer-facing AI platform capable of generating highly realistic videos in minutes.

As with most new technologies, there are bad actors who are eager to take advantage. Deepfake-powered scams have become more common across the internet, and the motivations behind these scams vary widely. Copyleaks recently highlighted the growing issue through research that examines scams involving celebrities like Taylor Swift and other public figures, highlighting how quickly manipulated media can spread online.

In recent years, several categories of deepfake scams have emerged as particularly common.

1. Celebrity

One of the most widespread forms of deepfake fraud involves celebrities. Scammers frequently use synthetic footage featuring public figures to promote fake products, investment schemes, cryptocurrency scams, or fraudulent giveaways. Because audiences already recognize and trust celebrities, these scams can appear highly convincing, especially when paired with realistic audio. Last year for example, several scammers were arrested in Brazil for creating ads with supermodel Gisele Bundchen advertising fake giveaways.

2. Romance

Romance scams and catfishing have existed for years, but deepfake capabilities have added a new layer of sophistication. AI-generated images, videos and voices have made it easier than ever to make people believe they’re in a relationship with a nonexistent person.

Bad actors take advantage of someone’s emotions by building a connection with their target who thinks they are real, and once they gain their trust, the scam typically escalates into requests for money or sensitive personal information. People have been scammed out of thousands of dollars and have even lost their houses to scammers preying on their emotions.

3. Political and Misinformation

Political races and global events have become a massive battleground in the fight against deepfakes. AI is often used to depict public figures making inflammatory statements they never made, often in the context of an opponent’s campaign video. As AI video tools improve, political deepfakes are becoming easier to produce and harder to identify quickly – especially when clips spread rapidly across social platforms before fact-checkers or news organizations can respond. There have been many attempts to enact laws to ban the use of deepfakes, especially on a local or state level. And while some have been successful, there is still a large grey area open for manipulation.

4. Loved One or Family Emergency

Using AI voice cloning and video generation tools, scammers can impersonate family members or loved ones during fake emergencies. Modern tools can create convincing audio or video messages that appear to come from a loved one who is in need of desperate help. Targets may receive frantic calls or videos claiming someone has been arrested, kidnapped, injured, or stranded and urgently needs money. These scams are especially heinous, yet effective because they prey on one’s emotional panic and urgency to pressure victims to act before verifying the situation – such as a daughter getting in a car accident or getting arrested. And as people share more public audio/visual content online than ever before, these scams are quite easy for bad actors to execute.

5. Fake Security Footage and Fabricated Evidence

Deepfake technology is also increasingly being used to create fake surveillance footage or manipulated recordings, some of which are even entered as evidence in court cases. They are convincing enough to falsely implicate an individual or an organization with the goal of extorting money or harming their reputation. Since security footage is often perceived as objective evidence, deepfakes can appear especially credible to viewers unfamiliar with the technology.

Tips for Spotting Deepfake Video Scams

As deepfake scams become more sophisticated, individuals and organizations need to be increasingly cautious when encountering suspicious video content online. One of the biggest red flags is unsolicited outreach from celebrities or public figures. Celebrities are not personally messaging users for promotions of giveaways, so one should not trust anything they receive that’s allegedly from someone famous.

There are also several visual and audio warning signs that may indicate manipulation, including:

• Unnatural facial movements
• Lip-sync mismatches
• Inconsistent blinking
• Distorted edges around faces or hands
• Robotic speech patterns
• Awkward pauses
• Audio that feels slightly out of sync with the video

Context is equally important. Before trusting or sharing suspicious content, verify where the video originated, whether reputable sources are reporting the same information, and whether the timing or claims make sense. As deepfake technology advances, human skepticism is equally as important as detection software, and the more people are aware of what to look for, the less effective deepfake scams will be.

FAQ