Detection of adulteration in milk is an important problem to overcome for the dairy industry. In this study, some physicochemical and microbiological properties of buffalo and cow milk, and the detection of adulteration by artificial intelligence algorithms using FTIR spectroscopy measurements obtained on 13 different concentrations (0.2-10 % (v/v)) buffalo-cow milk mixtures were investigated. The adulteration performances of six different artificial intelligence algorithms and the performances of SIMCA and DD-SIMCA from chemometric methods were analyzed. Additionally, it has been investigated to achieve high performance with fewer measurements by reducing the number of FTIR spectroscopy measurements with particle swarm optimization (PSO). As a result, it has been observed that FTIR and artificial intelligence-based algorithms provide significantly higher results. It was observed that the features selected with PSO reached a 90.38 % accuracy value in the Ensemble Bagged Tree algorithm. It's thought that the rapid results of artificial intelligence-aided systems will provide convenience to experts working in the field of food adulteration.