Improving the quality of education remains one of morocco’s major long term challenges especially given its impact on the country’s social an economic development. The PISA 2022 results offer a useful snapshot of how Moroccan 15 year old students are performing particularly in math where their scores still fall short of the OECD average. Rather than simply restating these findings this study seeks to understand what lies behind students performance and to explore whether these factors can be anticipated through the use of AI and ML techniques. Working with the extensive PISA dataset allows us to sort through a wide range of indicators related to students backgrounds, schools, socio-economic circumstances, cultural resources, access to technology and personal learning habits. When these variables are considered together a better picture emerges: mathematics achievement is not the result of a single influence but of the interaction between several dimensions such as school support, family cultural capital, digital access and the way students approach their learning. Machine learning models are particularly useful for identifying subtle links that traditional methods do not detect. The study also makes it possible to identify students who need support at an earlier stage and to inform data-driven education policies, with the aim of building a more equitable and effective mathematics education system in Morocco. This study adopts a machine-learning-oriented exploratory framework. The findings highlight predictive associations rather than causal relationships and should be interpreted in light of the methodological limitations related to the use of large-scale assessment data.