Research aims to defeat spam with AI and algorithms
Researchers received funding from the Global System for Mobile Communications Association to tackle the global problem of spam bots and SIM farms.
Every day, millions of people feel their phones buzz with the world’s most irritating notification: a spam call.
According to the U.S. Public Interest Research Group, there were 29.6 billion of these unwanted robocalls in 2025 alone. Spam, scam, or robocalls are calls made to mobile phones from either a program or a real person with the intent to steal information and money.
But a team of Virginia Tech researchers from the Brain-Inspired Computing, Communication, and Security Center based in the greater Washington, D.C. area is working to put an end to spam calls altogether.
“We’ve discovered a lot of these spam calls actually come from SIM farms plugged into the mobile networks,” said Yaling Yang, professor of electrical and computer engineering in the Institute for Advanced Computing. “They’re real SIM cards plugged into large setups that can place actual phone calls. So from the network’s point of view, everything looks legitimate. That’s why spam blockers often have a hard time catching them — it’s a real number, and it behaves like a real person making a call.”
SIM farms can be set up cheaply using low-cost specialized hardware, often called SIM boxes, along with inexpensive SIM cards. Installation is possible anywhere, in any country. But like the mythical Hydra, for every farm that comes down, two more sprout in its place.
Permanently stopping SIM farms and spam calls requires tools that haven’t been invented yet due to the one thing phone companies protect as much as possible: customer data and commercial secrets.
“That’s where our project comes in — it bridges that barrier,” said Yang. “We’re going to build a digital twin framework that can replicate the behaviors of a real mobile network, which researchers can use to develop real solutions to the spam problem.”
A digital twin is a digital model of a real-world object.
Supported by funding from the Global System for Mobile Communications Association, Yang and her research colleagues will not only share the digital twin framework with other researchers to develop algorithms to identify spam, but also create their own detection algorithms.
Using artificial intelligence, researchers can detect the differences between bot and human accounts, testing a wide range of indicators to consistently identify the bad actors.
Yang and David Simpson, professor of practice in the Department of Business Information Technology, presented initial findings of this work at the 2026 Mobile World Congress in Barcelona in March. Their scam bot roundtable helped attendees gain insights into how digital twin environments can model SIM farm operations and measure fraud detection accuracy.
“I’m most excited about the potential that we are going to solve a problem that will have a huge impact on the world,” said Yang. “Spam bothers everyone around the world, and what we’re building can fundamentally change ecosystems and turn these problems to a direction that is more positive.”
About the team
The grant is a collaborative effort between the Institute for Advanced Computing, Pamplin College of Business, and the Bradley Department of Electrical of and Computer Engineering. The research team includes:
- Yaling Yang, principal investigator
- Lingjia Liu, the Andrew J. Young Professor of electrical and computer engineering and co-director of Wireless@Virginia Tech
- Yang “Cindy” Yi, professor of electrical and computer engineering, Bradley Senior Faculty Fellow, and co-director of the Multifunctional Integrated Circuits and Systems Center
- David Simpson, professor of practice and retired rear admiral of the U.S. Navy
The Global System for Mobile Communications Association is the advocacy and lobbying organization for the mobile communications industry, representing more than 750 mobile operators as full members and 400 companies in the broader mobile ecosystem as associate members.
