Skip to content

Bangladeshi researcher builds AI Shield against cyber attacks on smart cars

Detection systems that anticipate adaptive cyberattacks may secure the future of self-driving cars

US-based Bangladeshi researcher Dilruba Shareen is developing artificial intelligence systems that detect cyberattacks on connected vehicle networks, including attacks specifically designed to evade detection.

Connected vehicles communicate through Vehicular Ad-Hoc Networks (VANETs), exchanging position, speed, and hazard data up to ten times per second to prevent collisions. The US National Highway Traffic Safety Administration (NHTSA) has estimated that vehicle-to-vehicle communication could help address up to 80 percent of crashes involving unimpaired drivers.

But this safety promise depends entirely on trust. A single vehicle broadcasting falsified data can trigger phantom braking, misdirect traffic, or cause collisions.

Shareen, a doctoral researcher in Computer Science at Oklahoma State University, targets the hardest version of this problem: reaction-aware insider attacks, in which adversaries observe how detection systems respond and adapt to evade them.

Her frameworks build a continuous behavioral record of every vehicle, much like a credit history. An attacker can easily fake a single message, but sustaining a false history over time becomes vastly harder. Her research has been recognized at the field’s flagship venues, IEEE GLOBECOM and the IEEE 104th Vehicular Technology Conference.

The implications are significant for the United States, where 36,640 people died in traffic crashes in 2025 despite four consecutive years of declining fatalities.

As federal and state agencies invest in vehicle-to-everything (V2X) deployment and autonomous vehicles expand across American roads, intelligent transportation systems are increasingly treated as critical infrastructure. Detection frameworks that remain reliable against adaptive attackers address a security gap that must be closed before nationwide deployment.

Her research is grounded in over 14 years of industry experience in data science and machine learning, including senior analytics leadership at Banglalink and Robi Axiata. Her work also carries forward-looking relevance for Bangladesh, where the World Health Organization estimates roughly 32,000 annual road deaths, and a recent government survey found that 66 people die from road traffic injuries every day, with economic losses estimated to exceed 1.5 percent of GDP.

While dense traffic and limited infrastructure put connected vehicle deployment years away, Bangladesh, as the nation that leapfrogged landlines to go mobile-first, can adopt intelligent transportation with security designed in from the outset.

Shareen’s research bridges the gap between the promise of autonomous mobility and the security needed to realize it by building vehicle networks that can withstand the adaptive adversaries of tomorrow.