The development of Computerized Adaptive Testing (CAT) is a key initiative in support of “Energy Efficiency and Environmental Quality” efforts. The use of CAT in assessments can reduce paper usage and printing, thus reducing energy consumption and environmental impacts associated with paper production and disposal. By utilizing an efficient CAT platform, testing can be conducted both offline and online, reducing the need for physical test transportation, which contributes to carbon emission reduction and enhances the efficiency of test administration. Meanwhile, in CAT development, Artificial Intelligence (AI) can be employed to manage and adapt tests based on participant responses, meaning that test questions are tailored to the individual’s abilities. This research employs the R program with the mirtCAT package. The study uses simulated participant test response data, demonstrating that the best item selection criterion and participant ability estimation method are Maximum Information (MI) criterion and Expected a Posteriori (EAP) method to make CAT usage more effective and efficient, ultimately conserving energy.