Generative AI is giving old-school scammers a terrifying glow-up in Georgia and South Florida, and the money is vanishing fast. Voice cloning, fake video endorsements and slickly spoofed identities are turning familiar cons like phony grandparent emergencies and bogus investment pitches into eerily convincing traps. Among the local victims: a Cobb County couple who say they were fleeced out of nearly $800,000 and a South Florida immigration attorney whose face and voice have been hijacked in hundreds of sham videos.
According to the FBI's 2025 Internet Crime Report, the Internet Crime Complaint Center logged 22,364 complaints in 2025 that referenced AI, with adjusted losses of about $893 million tied to those cases. Overall, the IC3 received more than 1 million internet-crime complaints last year, with reported losses topping $20 billion, a snapshot of how quickly digital fraud tactics are evolving. FBI analysts say this first-year breakout on AI-related crime likely underestimates the scope, because many victims still do not realize AI was part of the con.
State-level numbers show how close to home this is hitting. Reporting by CBS Atlanta notes that Florida logged roughly 350 AI-linked complaints and nearly $39.9 million in related losses last year, while Georgia recorded 71 complaints tied to AI and more than $10.4 million in losses. One case highlighted by authorities: a Cobb County news release describes a retired couple who lost about $800,000 in a sophisticated cryptocurrency scheme that used an AI-generated "ghost site" to make fake investments look real. Officials caution that even these eye-watering totals are probably low, since many people never report being scammed.
How AI Makes Old Scams Harder To Spot
Today’s generative tools can clone a voice from just a few seconds of audio, spoof caller ID and crank out realistic photos or video testimonials that give scammers instant credibility. The FBI’s analysis shows that investment schemes and so-called "distress" or grandparent scams rank among the costliest AI-enabled frauds, with investment complaints making up the largest share of AI-related losses. Because these systems make it easy to mass-produce tailored, highly specific messages, investigators say fraudsters are getting better at combining personalization and manufactured urgency to push victims into sending money before they have time to double-check.
South Florida Example: Likeness Used To Target Immigrants
In South Florida, immigration attorney Angel Leal told CBS Miami that scammers built hundreds of fake videos and profiles using his image and voice to promote bogus immigration services. Leal says his team has had to remove thousands of fraudulent social media accounts and that the impostors often steer would-be clients to WhatsApp numbers before demanding payment through Venmo, Zelle or similar apps. That push for quick, off-platform transfers is a big reason recovering the money is so difficult once it leaves traditional bank channels.
What Banks And State Offices Say You Should Do
Financial institutions are trying to get ahead of the wave. Truist warns that fraudsters are leaning on AI to create urgent, believable impersonations and urges customers to slow down, treat surprise payment requests with suspicion, verify identities using trusted phone numbers or contact methods and keep a close eye on account activity. For a broader checklist, the Georgia Attorney General's Consumer Protection Division highlights classic warning signs such as unsolicited outreach, pressure to act immediately and demands for hard-to-trace payment methods, and it directs victims to report suspected scams to state or federal agencies. Officials stress that a few simple verification steps, rather than speed, remain the strongest shield against AI-driven impersonation.
If you think you are being targeted, experts say to try to halt the transfer, contact your bank and local law enforcement, and consider filing a complaint with the FBI’s Internet Crime Complaint Center. Basic verification, such as calling a known number, checking directly with family members or agreeing on a private code word, is still one of the easiest ways to stop most AI-enabled scams before any money leaves your account.
