In urban areas, a significant increase in vehicles day-by-day leads to challenges like accidents and pollution such as air and noise. The emission of Carbon dioxide (CO2) from vehicles leads to air pollution. The major cause of increased emission is traffic conditions in urban areas. Thus, Quality of Service (QoS) becomes a very challenging research problem by considering eco-friendly and reliable transportation. The appropriate congestion or traffic situation management techniques reduce the possibilities of accidents and pollution. The congestion control methods should take into account the properties such as fairness, decentralization, network characteristics, and application demands in VANET. The current methods failed to address all such properties and trade-offs for VANET communications. This paper proposed Adaptive Congestion Aware Routing Protocol (ACARP) for VANET using the dynamical artificial intelligence (AI) technique. In ACARP, the adaptive congestion detection algorithm is designed using the type-2 fuzzy logic AI technique. The fuzzy model builds to detect the congestion around each vehicle using three fuzzy-inputs such as bandwidth occupation, link quality, and moving speed. Using three parameters, the fuzzy rules are designed in the first phase. In the second phase, the inferences model introduced where the fuzzy decision has been made. In the last phase, defuzzifuzification is applied and the congestion probability estimated in the range of 0-1 for each vehicle. Then the status of congestion detection updated using the pre-defined threshold value for each vehicle. The status of congestion and its probability values have been utilized to establish safe and reliable routes for data transmission. It saves significant communication overhead and hence CO2 emissions in the network. The simulation results prove that the planned protocol improved the QoS presentation and significant CO2 emission compared to underlying fuzzy-based methods.