Scammers Clone Daughter’s Voice To Extort $50,000 from Michigan Mom - 1

Image by Freepik

Scammers Clone Daughter’s Voice To Extort $50,000 from Michigan Mom

  • Written by Kiara Fabbri Former Tech News Writer
  • Fact-Checked by Sarah Frazier Former Content Manager

In a Rush? Here are the Quick Facts!

  • Scammers used her daughter’s voice to create a false emergency call.
  • The caller claimed to be part of a Mexican cartel.
  • Victims should verify identities before sending money to avoid scams.

A Grand Rapids mother nearly lost $50,000 over the weekend due to a sophisticated AI scam that utilized voice cloning technology, as first reported by Wood TV8 on Monday.

Mary Schat, whose daughter attends Hope College, experienced a terrifying incident on Sunday morning. While she was reading the news, she received a call from an unknown number in Holland.

“It didn’t say potential spam or anything. So, I thought, well, my daughter is in Holland,” Schat explained. “I better answer it. Maybe this is about her.”

Upon answering, she heard her daughter’s voice, but it sounded distressed. “I said, ‘Dori, you have to help me understand you. You need to calm down,’” Schat recalled. “What’s going on? I heard, ‘They’re taking me. They’re taking me.’”

Then a male voice intervened, claiming to be part of a Mexican cartel. He said her daughter had been in a car crash and demanded a ransom of $50,000 for her release, instructing Schat to meet at a nearby hardware store.

Panic set in as Schat thought about the situation. “When it’s your own daughter, I’m like, heck yeah. I’m going to give whatever they need. I was ready to go to the ATM machine,” she said.

While she negotiated with the caller, her husband contacted local dispatch and learned the call was likely a scam. He reached out to their daughter, who was safe at her apartment.

“It was definitely her voice. A mother knows her daughter’s voice,” Schat said, shocked by the replication of her daughter’s voice.

The Better Business Bureau (BBB) calls this scam voice cloning , enabled by software that can replicate a voice from a small audio sample. “I thought I was very aware of these hoaxes and these scams,” Schat said. “It was completely terrifying.”

To avoid such scams, the BBB advises against acting immediately or sending money, and recommends contacting the involved person first.

Meta Expands Llama AI Access To U.S. Government and Defense Partners - 2

Image by Alpha Photo, from Flickr

Meta Expands Llama AI Access To U.S. Government and Defense Partners

  • Written by Kiara Fabbri Former Tech News Writer
  • Fact-Checked by Sarah Frazier Former Content Manager

In a Rush? Here are the Quick Facts!

  • Meta is making its Llama AI model available to U.S. national security agencies.
  • Meta’s AI “acceptable use policy” restricts military use but allows U.S. defense access.
  • Llama access extends to Five Eyes countries: U.S., UK, Canada, Australia, New Zealand.

Meta has announced on Monday it will make its open-source AI model, Llama, available to U.S. government agencies, and contractors focused on national security.

Bloomberg noted that although Meta’s AI models are open source and freely available to developers, the company’s “acceptable use policy” prohibits their use in projects related to “military, warfare, nuclear industries or applications, [and] espionage,” among other areas.

However, Meta is making an exception for U.S. defense agencies and their contractors, aiming to expand its presence in the public sector and strengthen Llama’s role in the fast-growing AI arms race, as noted by Bloomberg.

These partnerships will enable government bodies to leverage Llama’s capabilities for a variety of applications, from defense operations to public service improvements.

Specific projects underscore Llama’s versatility in tackling complex, data-intensive tasks. For example, Oracle is integrating Llama to streamline aircraft maintenance, aiming to improve diagnostic precision and reduce repair times.

Scale AI is adapting Llama to aid national security teams in mission planning and assessing vulnerabilities. Lockheed Martin has embedded Llama within its AI Factory, supporting code generation and data analysis, which enhances both military and business processes.

Meta says that leading cloud service providers like AWS and Microsoft are hosting Llama in secure environments to safeguard sensitive data. IBM’s watsonx is another key player, offering Llama’s capabilities in self-managed data centers for defense applications.

Strategically, open-source AI is positioned to strengthen the U.S. in the global AI competition, as nations like China make substantial investments in similar technologies. Meta emphasized the critical need for the U.S. to lead in this rapidly advancing field.

The Verge points out that Chinese researchers recently used Meta’s Llama 2 model to develop an AI system for China’s military. In response, a Meta spokesperson told Reuters that using an outdated version of an American open-source model is insignificant, given China’s trillion-dollar investment to overtake the U.S. in AI.

A Meta spokesperson confirmed to Bloomberg that Llama is also being made available to government agencies and contractors in the UK, Canada, Australia, and New Zealand. These countries, along with the U.S., are part of the intelligence-sharing alliance known as Five Eyes.

Forbes points out that Meta’s open-source approach to AI sets it apart from competitors like OpenAI and Google, who prioritize closed ecosystems for commercial gain. By keeping Llama open-source, Meta enables greater customization and flexibility for enterprises.