A New Weapon Against Fraud
On July 8 at 2:00 PM, U.S. startup Savi Security launched an app that leverages artificial intelligence to combat fraud. The idea was born after the founders—brothers Patrick and Ryan Coughlin—saw their mother nearly fall victim to a phone scam demanding $1,200. Both brothers bring backgrounds in cybersecurity and tech, which helped shape the app’s development.
How It Works and What It Costs
The Savi Security app analyzes phone calls in real time to flag potential scams. Just three seconds of audio are enough to clone a person’s voice, highlighting the need for such protective technology. According to the U.S. Federal Trade Commission, American losses from online imposter scams could reach $3.5 billion in 2025—triple the amount lost in 2020.
The Savi team also built a free service called Scamwise, which received roughly 100,000 inquiries in four months. A subscription to the new app costs $8 per month or $63 per year, and one account can be shared across an entire household with no limit on users.
Patrick Coughlin, one of the co-founders, noted: 'Modern generative models have made these attacks much cheaper and more accessible.'
The launch of Savi Security reflects growing concern over digital fraud, especially as online threats continue to rise. Using AI to fight scams could be a key step in addressing these issues, given the potential financial losses tied to new criminal technologies. The strong demand for Scamwise also suggests a significant appetite for effective solutions, pointing to further growth in this market.
The rise of AI technologies not only aids in combating fraud but also poses new challenges, as evidenced by recent incidents like the deployment of ransomware by an AI agent. This alarming trend highlights the dual-edged nature of advancements in artificial intelligence, making it crucial for users to stay informed about both protective measures and emerging threats. For a deeper understanding of how these AI developments impact cybersecurity, read more about the independent deployment of ransomware by AI.