To spot a deepfake video or fake WhatsApp forward, look for unnatural eye blinking, mismatched lighting, robotic audio cadences, and language designed to trigger immediate panic or anger. Check if a message has the "Forwarded many times" tag, and always verify explosive claims with a secondary news source. Most importantly, never transfer money based on a video or voice note without speaking to the person live.
We have entered an era where seeing is no longer believing.
In India, the rapid advancement of Generative AI has collided with the world's largest messaging user base. The result is an epidemic of highly convincing fake content. Scammers use "deepfakes"—AI-generated audio and video—to impersonate politicians during elections, create fake celebrity endorsement videos for investment scams, and clone the voices of family members to extort money.
Simultaneously, traditional text-based misinformation continues to spread like wildfire across family WhatsApp groups, often inciting panic or even real-world violence.
Here is how you can train yourself to spot AI-generated scams and fake forwards before they cause harm.
How to Spot an AI Deepfake (Video and Audio)
Deepfake technology improves every month, but it still leaves digital artifacts. If you receive a video or voice note asking for money, making an explosive political claim, or offering a guaranteed investment, look for these signs:
1. The "Uncanny Valley" in the Eyes and Mouth
AI struggles to render realistic micro-expressions. In a deepfake video, watch the person's eyes and mouth closely.
- Do they blink naturally, or do they stare blankly for too long?
- Does the movement of their lips perfectly match the audio, especially on harsh consonant sounds (like P, B, and M)?
- Do their teeth look blurry or merged together?
2. Inconsistent Lighting and Shadows
AI often generates faces by stitching different images together. If the lighting on the person's face comes from the left, but the shadow on the wall falls to the left as well, the video is fake. Pay attention to the edges of the face and hair—deepfakes often have a slight blur or "halo" effect where the fake face meets the real background.
3. Robotic or Monotone Audio Clones
Audio deepfakes (voice cloning) are currently the most dangerous tool for financial scams. A scammer can take a 3-second clip of your family member's voice from Instagram and clone it to call you, claiming they are in an accident and need money. To spot an audio clone:
- Listen for strange pauses or a lack of natural breathing sounds.
- The voice may sound slightly metallic or robotic.
- The Ultimate Test: Ask a personal question that only the real person would know the answer to, or insist on switching to a live video call.
How to Spot Fake WhatsApp Forwards
Text-based misinformation is less technologically advanced but relies heavily on psychological manipulation.
1. The "Forwarded Many Times" Tag
WhatsApp introduced the double-arrow "Forwarded many times" label for a reason. If a message has this tag, it means it has passed through a chain of at least five different chats. It is highly unlikely to be original, verified information. Treat it with extreme skepticism.
2. Emotional Manipulation (Anger or Fear)
Fake news is designed to bypass your logical brain and trigger an emotional response. If a message makes you instantly furious, terrified, or demands urgent action (e.g., "Pass this to 10 people immediately or your account will be deleted"), it is almost certainly a hoax.
3. Vague Authorities
Misinformation often cites fake or vague authorities to sound credible. Phrases like "According to a top doctor at AIIMS" or "Forwarded from a senior police officer" are classic red flags. A real public service announcement will always have an official link to a government website (.gov.in).
4. The URL Trick
Scammers spread phishing links via forwards by slightly misspelling famous URLs. You might see amazon-diwali-gifts-free.com instead of amazon.in. Never click a link in a forwarded message without verifying the domain.
The Root Cause: The Anonymity Problem
Why do deepfakes and fake forwards spread so easily on platforms like WhatsApp, Telegram, and X (formerly Twitter)?
The answer is anonymity.
When anyone can create a fake profile or buy a burner SIM card, there is zero accountability for starting a rumor or sharing a deepfake. A scammer can create a fake "News Channel" group on Telegram, seed an AI-generated video of a bank collapse, and watch it spread to millions while remaining entirely hidden.
When you cannot trust the content (because AI can fake it), you must be able to trust the identity of the sender.
The AirlockChat Defense: Verified Identity
We built AirlockChat on the belief that accountable communication is the only defense against the AI misinformation era.
- No Burner Accounts: Every user on AirlockChat must verify their identity using their government ID via DigiLocker. If someone attempts to spread a malicious deepfake or run a financial scam on our platform, their verified legal name is attached to that action.
- Mutual Consent Stops Viral Spam: Fake forwards go viral because people blast them to hundreds of contacts and groups. On AirlockChat, you cannot message someone without their explicit consent. This structural friction prevents the rapid, viral spread of automated spam and panic-inducing forwards.
- Transparent Reporting: If a user is caught spreading malicious deepfakes, they can be reported. Confirmed reports result in citations on their profile, warning others.
Key Takeaways
Generative AI has made it incredibly cheap and easy for scammers to create convincing fake audio, video, and text. Protect yourself by looking for visual artifacts, robotic audio, and emotional manipulation. Never transfer money based on a digital message without verifying the person live. As AI makes content increasingly untrustworthy, the only solution is to use platforms like AirlockChat, where ID verification and mutual consent guarantee that the person sending the message is exactly who they claim to be.