If spam calls already pester you on a daily basis, brace yourself: It’s going to get a lot worse.

In fact, nearly half of all calls made to mobile phones in the United States will be fraudulent by 2019, according to a projection by the call protection company First Orion. The company projected this figure by analyzing more than 50 billion calls in the U.S. for over 18 months. It’s a scourge: 3.3 billion robocalls were made in April alone, with scammers making away with a staggering $8.9 billion in March, according to TrueCaller. Thankfully, researchers at Purdue University tell Inverse that they may have finally found a way to put a stop to the madness with a new app, CEIVE, that can protect users from the brunt of these spam calls.

“There have been solutions to this problem proposed in papers, but none have been deployed in the real world,” says Chunyi Peng, co-author of the research behind the app. “We want to build a solution which can protect users by detecting if the call is a spoofed caller ID or not.”

youmail spam call stats
There was an estimated 3.36 billion robocalls placed in April 2018, a 6.5% month-on-month increase in robocall volume, and a 34.4% increase vs. April of last year.

How CEIVE Detects Spam Calls

Other spam blockers exist, but these rely on pre-existing databases and “spam protection engines” to re-route your calls. CEIVE, short for callee-only inference and verification, is different, and uses a signaling protocol to identify imposters using the most common tools with perfect accuracy.

Every time someone gives you a ring, a call session begins and both the caller and receiver are said to be in an “in call” state. This is where CEIVE steps in, its app starts a separate, background session when someone calls you to check the state of the caller’s number. If they’re found to be “idle” or “dialing” and not “in call,” that number is being used as a disguise and CEIVE warns the user with a push notification.

Think about it like receiving a Twitter direct message from someone claiming to be your friend and realizing that their handle is different. CEIVE can check the signaling state of an incoming call using a machine learning technique to see if the actual number’s state is identical to the number being used in a session within seconds. It’s not fool-proof, but consumers are still hungry for any tools at all that can put a stop to the madness. Complaints to the Federal Trade Commission about spam calls have quintupled in the last few years.

“If our technology detects a spoof, we are very confident in our prediction,” Peng explains. “But if a false caller ID is not detected that does not mean a spoof is not present. There are certain advanced forms of attacks and certain cellular carriers that CEIVE isn’t compatible with.”

The app has shown perfect accuracy in all spam call cases that use popular fake caller ID tools. However, attacks that are able to manipulate the signaling state of the spoofed number can circumvent this defense tool.

CEIVE has also proven effective with four major U.S. cellular carriers: AT&T, T-Mobile, Sprint, and Verizon, but it won’t work with smaller carriers like MetroPCS, Boost Mobile, or Cricket.

The app is also currently unavailable for download because Peng and her two other team members don’t haven’t had the resources to make a user-facing interface. They are currently reaching out to Google to hopefully make this a reality.

A solution to the seemingly endless stream of spam calls could soon be in sight, at least for U.S. mobile users.