Savi launches app to block AI-generated kidnapping scams

Savi Security launched a consumer app for iPhone and Android on Tuesday that screens texts, voicemails and incoming calls for AI-driven scams, the company said. The startup raised $7 million in seed funding led by Acrew Capital, with participation from Magnify Ventures, TTCER, and Resolute Ventures.

What does Savi’s app do?

Savi’s app screens messages and calls in real time and includes a live-call monitoring option that lets the app add a live agent to listen during suspicious conversations. During a flagged phone call a user can opt to add the app’s live agent as a listener, and Savi watches for behavioral tells that indicate a grift while the call is in progress. The paid plan costs $8 per month or $63 per year, covers an entire family with no cap on the number of users, and can be used across kids, spouses and older relatives.

How was the AI trained and built?

Savi trained its scam-detection model using submissions to a free website the founders launched, Scam Wise, which accepts suspicious texts, photos and emails without registration. Scam Wise has received 50,000 submissions since it launched about four months ago and now grows by about 10,000 submissions or more each week, the founders said. The startup is currently mostly using Google’s Gemini, and has built its software on an AI gateway so it can tap other models as needed, including voice detection–specific options.

What prompted the startup and how big is the problem?

The company was founded after a near-miss two years ago when the founders’ mother received a phone call that spoofed the daughter’s number and used an AI-cloned voice to claim the daughter had been kidnapped. The scammer said, "If you don't pay us $1,200 right now, we're going to kill your daughter in the parking lot of the local Walmart," according to the founder’s account. The founders point to cheap, widely available large language models and audio-cloning tools as the reason such attacks are now feasible and profitable at scale.

Regulators and security vendors say the problem has escalated. The FTC said last month that people reporting online crimes collectively lost $3.5 billion to imposter scams in 2025, triple the amount in 2020. Research from 2025 by Malwarebytes found Gen Z is targeted more often with text scams than other generations and fell for them about 25 percent of the time.

Why it matters

AI tools have driven down the technical and financial barriers to producing realistic voice and text scams, shifting sophisticated deception from enterprise targets to everyday consumers. A product that can screen incoming communications, spot contextual red flags and intercede during live calls tackles the window where many victims must decide under pressure, especially older Americans and younger people targeted over text. The combination of consumer-priced subscriptions and family coverage aims to protect networks of vulnerable contacts rather than single devices.

What to watch

Watch whether Scam Wise’s submission growth sustains beyond the current roughly 10,000 new reports per week and how Savi expands beyond Gemini by integrating other voice-detection models available through its AI gateway. Also monitor user uptake of the live-call monitoring feature and whether regulators update guidance or reporting after the FTC’s $3.5 billion figure for 2025.