Globally, machine learning technologies are advancing quickly and having an impact on a wide range of fields and sectors including education. To educate the students on the opportunities it presents, a matching inclusion in the educational curricula is required. One area of growing importance for educational innovation and development is the use of machine learning in science education curricula at higher learning institutions. The study explored the integration of machine learning in science education programmes in universities in South Africa and Nigeria. The study adopted a descriptive survey research design with a sample size of 160 (120 students & 40 lecturers) which were determined using simple random sampling techniques. The instrument used for the study was a questionnaire which was validated by three experts and a reliability index of 0.88 was obtained using Cronbach alpha. Mean, standard deviation and independent samples t-test were used to analyse the data. The results of the findings revealed that the extent of integration of machine learning into science education curricula as perceived by both students and lecturers is low. Also, the influence of machine learning on students' learning outcomes includes improved problem-solving abilities, enhanced critical thinking skills, and understanding of complex concepts while the challenges faced by institutions in incorporating machine learning into science education programs include a lack of adequate training for educators, lack of resources and high cost of machine learning tools. The study recommended that institutions collaborate with industrial experts to develop a curriculum that incorporates practical applications of machine learning in the learning process.

Keywords: Machine learning, science education, perceptions, challenges, educational implications, learning outcome, universities.