Savi’s App Seeks to Protect Consumers from Persuasive AI Scams, Including Kidnap Ransom Threats

Brothers Patrick and Ryan Coughlin, both seasoned professionals in the tech industry (Patrick has experience in national cyber defense, as well as roles at Splunk and Cisco, while Ryan has focused on consumer products at Apple and Spotify), have launched an innovative security startup.

Savi Security seeks to protect everyday users from the rising threat of sophisticated AI-generated scams delivered via texts, emails, or phone calls.

The company recently raised $7 million in seed funding and is set to debut its app for both iPhone and Android this Tuesday. Acrew Capital led this funding round, with support from Magnify Ventures, TTCER, and Resolute Ventures.

The inspiration behind the startup originated from an alarming incident involving the founders’ mother.

Approximately two years ago, Patrick’s mother reached out to him, upset, stating that a man had contacted her claiming to have kidnapped Patrick’s sister. At that time, he served as the senior vice president of security products at Cisco, having joined there following Splunk’s acquisition of his cloud security firm TruSTAR for roughly $82 million in May 2021. Cisco later acquired Splunk in 2024.

During the conversation, her mobile phone displayed the caller ID of her daughter, Coughlin recalled. At that moment, “she thinks she hears my sister’s voice saying, ‘Mom, they’ve got me.’ Then there’s a horrifying scream, and my sister says, ‘You need to do what they say.’ A man then tells her, ‘If you don’t send us $1,200 right now, we will kill your daughter in the Walmart parking lot,’” he explained.

The scammer had successfully spoofed Coughlin’s sister’s phone number, convincingly imitated her voice, and mentioned the nearby Walmart she frequently visited.

Fortunately, his mother kept her cool, called her daughter, and confirmed she was safe. The alleged kidnapping was merely an AI-generated scam.

Coughlin indicated that he, like his mother, was profoundly unsettled.

“After reassuring my mom, I found myself contemplating: What has shifted in the realm of cybercrime that we’re now observing the same level of sophistication aimed at government agencies and Fortune 500 companies, now targeting consumers?”

The answer lies in affordable and powerful LLMs and generative AI technologies.

Before AI became prevalent, executing such scams against consumers wasn’t economically feasible; it demanded extensive target research and technology for voice spoofing, making it largely focused on affluent entities like corporations or governments, along with their necessary defensive systems.

“Currently, AI is making it easier for cybercriminals to exploit consumers,” Coughlin emphasizes. The financial barriers for executing these scams have significantly lowered, and relevant resources are readily available.

“You can mimic a voice with just three seconds of audio from a public social media post. We all leave digital traces behind—whether it’s casual chats or comments during a child’s sports game shared on Facebook.”

The FTC reported last month that individuals lodged complaints regarding online crimes leading to total losses of $3.5 billion due to imposter scams in 2025, a threefold increase since 2020. While older Americans are primarily affected, some studies indicate that Gen Z is also significantly susceptible. Research from 2025 by Malwarebytes revealed that Gen Z was targeted more frequently by text-based scams than other age groups, with a victimization success rate of 25%.

The Coughlin brothers conceptualized a real-time intervention tool to combat these threats.

They validated their idea and the AI scam detection model by establishing a free website called Scam Wise. This platform requires no registration and permits users to upload any dubious texts, images, or emails for assessment.

“We launched it about four months ago, and we’ve already received 50,000 submissions, which is now increasing by approximately 10,000 weekly,” Coughlin remarked.

Scam Wise has proved to be a valuable source of real-world data, assisting in refining Savi’s scam-detection AI model. The startup primarily employs Google’s Gemini but has developed its software on an AI gateway, enabling access to other AI models when necessary, including specialized options for voice detection.

This Tuesday, Savi introduced its paid offering: an app for iOS and Android designed for consumers that can screen texts, voicemails, and incoming calls for scams.

While similar features exist in various products (like Malwarebytes), Savi’s unique offering lies in its live-call monitoring function.

During a suspicious phone call, users can opt to add a live agent from the app as a listener. Savi actively monitors for behavioral signs that might indicate fraudulent activity throughout the call.

Savi’s pricing model is also distinctive. It costs $8 per month, reduced to $63 per annum, covering an entire family with no user limit. This allows one subscription to encompass the user’s children, spouse, parents, and any relatives needing tech support.

Coughlin pointed out, “AI has transformed the landscape regarding how accessible fraud has become. We’re unintentionally fostering fraudsters by lowering the barriers to deceiving individuals. We’re now not only facing organized crime syndicates but also everyday people who might be lured into committing fraud.”

Savi Security signifies a new era of anti-virus-like software, leveraging AI in real-time, akin to its malicious counterparts.