Por Renan Filho · 2026-10-07**Lead**
A surge in artificial intelligence–powered scams is prompting federal agencies and cybersecurity experts to issue a counterintuitive warning: the most effective shield against these hyper-realistic frauds is not a cutting-edge app or algorithm, but the same low-tech skepticism that has protected consumers for decades. From voice-cloning phone calls that mimic a grandchild in distress to AI-generated video messages that appear to come from a CEO, the tools available to criminals have never been more convincing — yet the fundamental defense remains a simple pause, a second look, and a direct verification call.
**Context**
The Federal Trade Commission (FTC) has reported a sharp increase in complaints involving generative AI, particularly in impersonation scams and phishing attempts that use synthetic voices or deepfake videos. In a consumer alert published earlier this year, the FTC noted that scammers are now able to scrape a few seconds of audio from social media to clone a person’s voice, then use that recording to call a family member and plead for emergency money. The agency’s advice is deliberately low-tech: hang up and call the person back on a number you know is real.
Similarly, the FBI’s Internet Crime Complaint Center (IC3) has tracked a rise in business email compromise schemes that incorporate AI-generated text to mimic the writing style of executives. In one widely cited case, a finance worker at a multinational firm was tricked into transferring millions of dollars after receiving a video call that appeared to show the company’s chief financial officer — a deepfake created from publicly available footage. The worker later told investigators that the request seemed urgent and the face on screen looked exactly like the CFO, but he had not verified through a separate channel.
These incidents are not isolated. AARP’s Fraud Watch Network has also warned that older Americans, who already lose billions annually to traditional scams, are particularly vulnerable to AI-enhanced fraud because the technology erases many of the red flags — bad grammar, unnatural pauses, pixelated images — that once gave victims a chance to recognize deception. The organization advises families to establish a simple code word that can be used to confirm identity during any unexpected request for money or personal information.
**Analise**
The emphasis on low-tech skepticism reflects a growing recognition among cybersecurity professionals that no software can fully protect a human who is emotionally manipulated or rushed into a decision. “The best firewall is between your ears,” is a phrase often repeated in training sessions, but it takes on new urgency as AI makes scams nearly indistinguishable from reality. Experts point out that scammers exploit the same psychological triggers — urgency, authority, fear, and love — regardless of the technology they use. The solution, therefore, is not to try to outsmart the AI but to change the behavior that scammers target.
This approach has implications for policy as well. Lawmakers in several states have proposed bills requiring AI-generated content to be labeled, but enforcement remains difficult, and many scams originate overseas. Meanwhile, tech companies are racing to develop detection tools, but those tools often lag behind the criminals’ innovations. The FTC’s message is pragmatic: assume that any unsolicited communication could be fake, and make verification a habit rather than an exception. That means calling the person back on a known number, asking a question only the real person would know, or simply waiting 24 hours before acting on any urgent request.
**Fontes**
This article draws on publicly available consumer alerts and reports from the Federal Trade Commission (FTC), the FBI’s Internet Crime Complaint Center (IC3), and the AARP Fraud Watch Network. Additional context is based on statements from cybersecurity researchers and law enforcement officials cited in major U.S. news outlets, including The New York Times and The Washington Post, covering the rise of AI-enabled fraud in 2025 and 2026.