On Tuesday, tech startup Savi Security launched an innovative iOS and Android app designed to protect consumers from highly convincing, AI-generated phone, text, and email scams, following a $7 million (£5.5 million) seed funding round. Founded by industry veterans Patrick Coughlin (former national cyber defence expert at Splunk and Cisco) and Ryan Coughlin (formerly of Apple and Spotify), the platform aims to neutralise sophisticated threats like simulated family kidnappings before victims part with their money.
The seed funding round was led by Acrew Capital, with additional participation from Magnify Ventures, TTCER, and Resolute Ventures.
A Terrifying Real-World Inspiration
The driving force behind the startup was a deeply distressing incident involving the founders’ own mother. Approximately two years ago, she received a phone call displaying her daughter’s caller ID. Upon answering, she heard what sounded exactly like her daughter screaming, “Mom, they’ve got me,” followed by a man claiming he would kill her in a local Walmart car park unless he received an immediate payment of $1,200.
At the time, Patrick Coughlin was serving as the Senior Vice President of security products at Cisco—a role he assumed after Splunk acquired his cloud security startup, TruSTAR, for a reported $82 million in May 2021 (Cisco subsequently acquired Splunk in 2024). Despite his professional background in high-level cybersecurity, the threat hit incredibly close to home.
The fraudster had successfully spoofed the sister’s phone number, cloned her voice, and referenced her frequently visited local supermarket. Fortunately, their mother kept her composure, hung up to call her daughter directly, and confirmed she was entirely safe. The entire kidnapping scenario had been fabricated using generative AI.
The Democratisation of Cybercrime
The incident forced Coughlin to reflect on how the cybercriminal economy has fundamentally shifted. Advanced tactics once reserved for targeting government agencies and Fortune 500 corporations are now being deployed against ordinary individuals. The catalyst for this shift is the widespread availability of cheap, highly capable Large Language Models (LLMs) and voice-cloning tools.
Historically, executing such targeted scams required extensive manual research and expensive technology, making them financially unviable against individual consumers. Today, bad actors can clone a person’s voice using as little as three seconds of audio extracted from a public social media video, such as a clip of someone narrating a child’s football match on Facebook.
The scale of the problem is rapidly expanding. The US Federal Trade Commission (FTC) said that reported losses to imposter scams reached $3.5 billion in 2025, representing a threefold increase since 2020. While elderly citizens remain highly targeted, younger demographics are also vulnerable. A 2025 study by cybersecurity firm Malwarebytes revealed that Gen Z is targeted by text-based scams more frequently than older generations, and users within this demographic fell for them about 25% of the time.
How Savi Protects Users in Real Time
To counter these threats, the Coughlin brothers built a real-time intervention engine. They initially tested their detection algorithms by launching Scamwise, a free, anonymous web tool where users can upload suspicious texts, emails, or images for instant verification. Within four months, the platform processed 100,000 submissions, with volume currently growing by over 10,000 queries per week.
This massive repository of real-world threat intelligence was used to train Savi’s proprietary detection models. While the startup currently relies on Google’s Gemini model, its software architecture utilizes an AI gateway, allowing it to seamlessly integrate other specialised models, including dedicated voice-analysis tools.
While standard spam filters exist across various security products, Savi distinguishes itself with its live-call monitoring capability. If a user receives a suspicious phone call, they can invite Savi’s live agent into the call. The software analyses the ongoing conversation in real time, scanning for specific behavioural patterns and linguistic indicators associated with fraudulent schemes.
Affordable Protection for the Whole Family
Savi has structured its pricing model to encourage household adoption. For $8 per month (or an annual subscription of $63), users can secure an unlimited number of family members under a single account. This allows a primary user to manage security protocols for children, spouses, and elderly relatives under one plan.
AI has fundamentally lowered the barrier to entry for digital fraud, effectively turning casual bad actors into sophisticated scammers. Savi Security intends to act as a modern equivalent to traditional antivirus software, deploying defensive AI in real time to protect everyday users from increasingly complex social engineering tactics.