Deepfake scams often look highly realistic. Scammers create fake videos, audio clips, or images using AI that appear genuine, making it easy to fool people. Fraudsters even replicate the voices of trusted figures such as bank officials, law enforcement, or family members to gain the victim’s trust. Today’s scammers are so clever that it can be very difficult to distinguish between real and fake content. Earlier, deepfake scammers mostly created funny celebrity videos, but now these scams have become more widespread, especially on social media platforms like Instagram and Facebook. Many reported cases show that victims have been trapped in illegal activities and suffered from financial fraud, misinformation, sextortion, and identity theft.
Types of Deepfake Scams
Nowadays, AI models can analyse public photos or voice recordings of an individual from any social media platform like YouTube, Instagram, and Facebook. Then these AIs create fake voices, images, and videos, undoubtedly well. An iProov survey found that 43% of people admitted they couldn’t tell if a deepfake was real or fake. Meanwhile, only one-third weren't even aware of these types of scams.
Below are some famous scams to raise awareness about deepfake scams:
Celebrity Deepfakes
This is one of the most popular scams, where scammers use deepfakes of celebrities to create fake videos that make someone appear or sound real. Recently, Tom Cruise’s deepfakes have gone viral on TikTok and Instagram reels.
In the name of AI, content creators have become increasingly advanced, with users creating explicit images of celebrities and manipulating viewers for the sake of publicity and to become famous. It brings reputational damage to the celebrities, causing embarrassment.
Financial Scams
Scammers replicate the voice of CEOs and bank officials to authorize fraudulent wire transfers or create fake investment opportunities using AI technologies. These scams target famous corporate leaders into sending large amounts of money to fraudulent accounts.
Impersonation Scams
Mimicking family members in distress, like a grandchild, to solicit emergency funds. The scammer asks for money for emergencies, travel expenses, or to help with an investment opportunity. They trick you by saying you’re the only one who can help, while isolating you from friends and family.
Romance Scams
People create convincing fake profiles on dating apps using deepfake images and voices to manipulate victims emotionally and financially. They may use deepfake technology to make themselves appear in video calls, making the relationship seem totally real. These scammers build special connections and get invested in the relationship so that they request financial help. If the person claims to live abroad or far away, it can be a red flag.
Phishing and Social Engineering Scams
Deepfakes of colleagues and the CFO were used to trick a finance worker into transferring funds. Modern AI can create live, real-time deepfake video calls, allowing criminals to attend authentic meetings and pressure victims. They fall victim to clicking on malicious links.
How Can You Spot a Deepfake?
In this digital age, deepfakes are difficult to detect, but there are still some points, or we can say weaknesses, that you can keep in mind while spotting the unrealistic ones. Here are some audio-visual inconsistencies you must consider:
Eyes and Blinking
One of the most common weaknesses of AI is that it fails to make natural eye movements. Check for unnatural blinking or eyes that seem frozen. But you must note that these gaps are being filled in new versions of AI.
Mismatched Facial Expressions
AI algorithms sometimes struggle to create an exact replica of the human facial features that doesn’t align with the emotions shown within the video. For example, you’re seeing a funny video, but the emotions of a person mimicking it aren’t matching the joyous ones, or the lips get stuck.
Unusual Facial and Body Alignment
Look for awkward head and body movements, jerky or unusual body positioning, or a face that seems not quite connected to the body. Check fine details like teeth, hair, and skin tone.
Hair
Fake images or videos often struggle to capture natural details such as frizzy, messy, or uneven hair. These imperfections make it easier to identify deepfakes, as AI still lacks precision in replicating fine hair textures.
Lip-Syncing Issues
If you notice even a slight discrepancy in lip-syncing, it's likely a sign of a deepfake. Sometimes it’s hard to identify, because AI can mimic so perfectly that you get confused, but you can spot it when the person turns their head or doesn’t pronounce something well-matched with the mouth movements.
Shadows, Lighting, and Skin Texture
Despite advancements in AI technology, deepfakes signal inconsistencies in skin textures, disturbing noise, sound, or lighting. Check for the reflections in glasses or other reflective surfaces. If you notice the face lighting is disconnected from the background, it could be a sign of a deepfake. AI may show incorrect glare, shadows, fuzzy details, or angles that don’t fit well with the physical scene.
Note: The overall quality and direction of light should be consistent across the entire video, not just the face.
Final Thoughts
To protect yourself from such deepfake scams, be cautious of unexpected videos or audios that involve the urgency of money or requesting personal information. Investigate the source of the information and cross-check it with trusted sources before you share it with others. Look for the traits of deepfakes mentioned above to verify the content’s authenticity.
Frequently Asked Questions
How Can I Identify Deepfakes?
AI sometimes removes a mustache or adds a beard, which is a clear sign that deepfakes may struggle with generating natural facial features and hair. These are some common signs that indicate deepfakes: Check for facial features like facial moles, blinking eye, mismatched lip-sync, incorrect body posture, stuck or moving eyes, shadows, incorrect lighting without matching the background, and different skin textures.
Which States Banned Deepfakes?
At least 10 states, including California, Florida, New York, South Dakota, Georgia, Virginia, Texas, Hawaii, Minnesota, and Illinois, have restrictions on creating deepfakes, and offenders face strict penalties, such as major fines and even long-term imprisonment. It was mainly implemented after the famous singer Taylor Swift faced exploitation because of her digitally spread pornographic visuals across many social media platforms like Twitter, Instagram, and more.
Does Google Block Deepfakes?
Deepfakes containing explicit content, like nudes or other sensitive content like pornography, within the ads will be banned under the policies of Google. Google bans explicit deepfake content, such as pornography, under its Ads and Content policies. One must be careful while creating deepfakes; it must not hurt someone’s sentiments, and one must be sure of the ethical concerns of an individual.
