Skip to content Search
With support of game theory, AI can help protect critical infrastructure from various attacks. Read the article by IDEAS NCBR experts from the "AI for Security" research team.

One of the pioneering applications of artificial intelligence, game theory and optimization techniques to protect critical infrastructure was the ARMOR system used at LAX airport in Los Angeles.

The challenge in protecting this airport, one of the largest in the world, is that “there are only a handful of canines of particular expertise, such as explosive detection ones. It means that it is simply impossible to provide constant coverage of a single explosive detection canine patrol per each terminal at LAX.” After the ARMOR system was introduced, overall crime detection at LAX increased sixfold, and drug smuggling detection increased tenfold.

AI offers great opportunities where human cognitive abilities are limited. “(…) By using game theory, we can develop a strategy to efficiently distribute scarce security resources for infrastructure protection.”

“(…) At the IDEAS NCBR research institute, our team ‘AI for Security’ builds Stackelberg models for various types of critical infrastructure. Currently, we have been focusing on developing software for protecting ports, LNG terminals, railways, and power grids.”

“A recent real-world example of the tactics that are explicitly modelled in our two-phase game are the actions of Lukashenko’s regime in Belarus which exploits immigrants to probe the border with Ukraine. (…). the Belarusian border guards are actively organizing and coordinating groups of immigrants in an attempt to breach it. Their objective is to expose and disrupt the Ukrainian defenses, which are obligated to respond to such attempts due to the threat posed by Russian saboteurs. (…) the border is not entirely impenetrable. Nevertheless, even in cases where a specific section of the border is unguarded at the moment of entry, the Ukrainian headquarters promptly dispatch a team to the area.”

Read the article by Dr. Tomasz Michalak (University of Warsaw, leader of the “AI for Security” research team at IDEAS NCBR), Dr. Michał Godziszewski (UW, IDEAS NCBR) and Dr. Andrzej Nagórko (UW, IDEAS NCBR), which was published in Poland’s Internal Security Agency (ABW) journal Terroryzm. Studia, analizy, prewencja [Terrorism. Studies, analyses, prevention]. Researchers describe how to use AI based on game theory to optimize critical infrastructure protection resources, specifically the Stackelberg model.

Read the article in English:

Featured news

Breakthroughs require freedom and patience
Krzysztof Walas joins Adra Board of Directors
I was told that doing computer science is not for women. Alicja Ziarko talks women in AI research