Sebaraya Tech – A deepfake is a photo, video, or audio clip created or altered with AI to show someone saying or doing something they never did. US security agencies, including the NSA, FBI, and CISA, warn that these fakes are now cheap to make with tools that run on an ordinary laptop.
You can’t rely on spotting every fake by eye anymore. What works better is a mix of visual checks, a quick look at where the content came from, and habits that make impersonation scams fail even when the fake is convincing.
How Scammers Use Deepfakes
The FBI describes several patterns: AI-made profile photos and fake IDs, images of celebrities “endorsing” products, and fake disaster photos used to collect bogus donations.
Voice cloning is used to imitate a relative in a staged emergency and demand money. Criminals have also posed as company executives in live video calls.
US security agencies add that deepfakes are used to impersonate leaders and financial officers to commit fraud. The common thread is pressure to act fast before you check.
Visual Clues in Images and Video
The FBI lists details that often give fakes away:
- Distorted hands or feet
- Unrealistic teeth or eyes, or faces that look blurry or irregular
- Odd accessories, such as glasses or jewelry that don’t sit right
- Shadows that fall the wrong way, or stray watermarks
- In video, lag, mismatched voice and lip movement, or unnatural motion
Treat these as warning signs, not proof. As tools improve, fakes may show fewer of these flaws, and a real photo can look strange too.
Check Where an Image Came From
Context often reveals more than pixels. Google‘s About this image tool shows when Google first saw a similar image, how other sites describe it, and sometimes metadata saying whether it was made by a camera or by AI.
To use it, open the image in Google’s image viewer, tap the info icon, choose About, then Learn more about an image. An image claiming to show yesterday’s event that Google first indexed years ago is a strong red flag.
The tool has limits. It’s available only in some regions, and Google notes that editing tools can change metadata, so it may be wrong. Confirm with other sources.
Defend Against Voice and Video Calls
Agree on a secret word or phrase with your family. The FBI recommends this so a caller claiming to be a relative can prove who they are.
Listen to tone and word choice; a cloned voice may not sound quite like the person you know. If a call asks for money or sensitive details, hang up and call back on a number you already have or one you looked up yourself.
The same rule works for a “boss” on a video call asking for an urgent transfer: confirm through a separate channel before acting. US agencies recommend real-time verification for exactly this situation.
Reduce What Can Be Copied
Deepfakes need source material. The FBI suggests limiting public posts of your voice and face, making social media accounts private, and keeping followers to people you know.
Never send money, gift cards, or cryptocurrency to someone you’ve only met online or over the phone.
If You Think You Were Targeted
In the US, report AI-enabled fraud to the FBI’s Internet Crime Complaint Center at ic3.gov, including payment details and how the contact began. Elsewhere, report to your national police cybercrime unit.
If money already left your account, call your bank immediately. The sooner they know, the better the chance of stopping the transfer.
Set up that family code word this week. It costs nothing and still works when the fake voice sounds convincing.
Komentar Ditutup! Anda tidak dapat mengirimkan komentar pada artikel ini.