Deborah Del Mastro transferred more than $5,000 after scammers used an AI-cloned version of her daughter’s voice in a kidnapping hoax. “It’s my daughter’s voice,” the California mother recalled. “And she’s having a panic attack.” Ordinary families are the ones these tools target. The technology spreads faster than detection systems can reliably track it.
When the Voice Sounds Real
Cybersecurity expert Morgan Wright said cybercriminals most often target ordinary Americans by exploiting personal vulnerabilities for quick payouts. Generative AI can remove language barriers, write convincing phishing texts and create realistic voice and video clones. A few seconds of audio, often taken from social media, can imitate a family member.
Del Mastro believed her daughter faced immediate danger and sent the money before she realized the call was a scam. Wright described the method as “manipulation, influence and deceit,” adding that urgency can make people lower their guard and act. The technology turns intimate trust into an opening for theft.
Wright said skepticism, rather than sophisticated detection software, is the best defense. He has trained federal, state and local law enforcement in behavioral analysis, interviewing and interrogation, including teams at the FBI and the National Security Agency. When a suspicious call or message comes in, he advises people to pause before responding.
“We’re not storming the beaches at Normandy; you’re not going to get arrested,” Wright said. “No law enforcement agency sends you an unsolicited text saying you’re about to get arrested.” The person receiving the message has to act: stop, question the pressure, and don’t respond in panic.
Scammers keep finding ways around updated security features, Wright said, leaving software detection as a moving target. “We hear about all the different models from xAI, Anthropic, OpenAI and Gemini. They change models constantly,” he said. “So, when the model changes, how effective is [detection software]? We don’t know because this stuff is moving so fast. What you see today will be drastically different tomorrow.”
Public Wants Rules; Power Keeps Building
A Reuters/Ipsos poll found that most U.S. voters believe President Donald Trump’s administration and Congress aren’t taking AI risks seriously enough, and want the government to write new regulations. The six-day survey ended Monday. Fifty-seven percent of registered voters said the Trump administration hadn’t taken the risks seriously enough or at all; 54% said the same about Congress. A third of self-identified Republicans shared that view.
The poll found that 84% of voters saw AI as a threat to American workers. Sixty-two percent said AI could get out of control and risk humanity’s future. Trump has repeatedly supported AI development and data-center construction, arguing they’re vital to U.S. competitiveness. He has also said the U.S. technology industry could fall behind Chinese counterparts if his administration doesn’t limit the technology’s use and development, an argument AI safety advocates call misguided.
A bipartisan majority backed stricter government regulation of AI: 80% of Democrats and 61% of Republicans. Fifty-six percent supported limits on data-center construction. The poll also found that 57% said the U.S. government shouldn’t use AI to choose targets for military strikes.
Daniel Kokotajlo, a former OpenAI staffer who quit in 2024 because he believed the company wasn’t behaving responsibly, said job losses and threats to human health and safety concern most Americans more than competition with China. A White House official pointed to Trump’s new task force, saying it is responsible for ensuring “America continues to lead the world” in the technology. The official said the administration’s priority is a booming economy, with workers experiencing limited volatility from short-term disruptions.
The Cost Lands Below
The poll ran online nationwide and gathered responses from 4,506 U.S. adults, including 3,526 registered voters. Its margin of error was 2 percentage points. Voters are asking for rules, limits and protection from risks to jobs, health and safety. The administration’s stated priority remains technological leadership and economic growth, while families like Del Mastro’s face the immediate cost when AI-enabled deception reaches their phones.