Updated

Updated · Kotaku · Aug 22

AI Scammers Ramp Up Deepfake Calls and Videos as Experts Warn Barriers to Entry Are Near 0

Updated

Updated · Kotaku · Aug 22

AI Scammers Ramp Up Deepfake Calls and Videos as Experts Warn Barriers to Entry Are Near 0

Q: If human senses can no longer detect AI voice clones, what secret low-tech method can save your family from financial ruin?

A shared family code word—or short passphrase—combined with an out-of-band callback is the best low-tech defense. Experts say human hearing is no longer a reliable filter. High-quality AI voice clones can be built from just seconds of audio, often scraped from social media or voicemail, and paired with caller ID spoofing to make a fake emergency sound real. That is why researchers, the FBI, and deepfake specialists increasingly recommend procedural checks over intuition. The practical fix is simple: agree on a secret word or phrase known only to close family members, and use it whenever someone makes an urgent request for money, secrecy, or personal information. If the caller cannot provide it, treat the call as fraudulent. Even if they do, the safer move is to hang up and verify through a separate trusted channel, such as calling a known number, texting another relative, or checking with someone who knows the person’s location. This advice is grounded in real cases. A Ferrari executive reportedly stopped an impersonation attempt by asking a personal verification question. Other victims were not so lucky: families and companies have lost thousands to millions of dollars after trusting what sounded or looked like a loved one or executive in distress. The key is not better ears, but better habits. Pick a code word that is memorable, private, and not discoverable online, share it only with trusted relatives, and rehearse using it before a crisis call comes.

3 articles · Updated · Kotaku · Aug 22

Deepfake video and AI voice-call scams are spreading more widely and becoming harder for average people to spot, with less tech-savvy people seen as most vulnerable.
MIT-linked AI Incident Database researcher Simon Mylius said fake content can now be produced by “pretty much anybody,” because the tools have become accessible with effectively no barrier to entry.
Doppel executive Bobby Ford said scammers are already increasing output because deepfakes that once took significant time and resources can now be made much faster and at higher volume.
Researchers and scam-watchers such as Kitboga and Jim Browning have exposed both AI call scams and deepfake video fraud, underscoring how quickly consumer-facing deception is evolving.