The swift advancement of multilingual artificial intelligence (AI) carries considerable consequences for decision-making across a range of fields, such as education, healthcare, finance, and governance. Systems of multilingual AI that can understand and generate some language naturally tend to reduce the language barriers to cross-linguistic communication, hence contributing to the reduction of such language barriers. This study focuses on the influence of multilingual AI on decision-making and capacity building for risk assessment in communication systems. The research shall be based on a qualitative methodology using descriptive and analytical techniques through a review of the available literature on multilingual AI and decision-making theories. Results suggest that multilingual AI systems could have effects on decision outcomes, user involvement, and cognitive functions in complex linguistic and cultural contexts. The study also raises issues on biases, decision accuracy, and non-representation of low-resource languages. These results contribute to the initiation of actions addressing societal implications of multilingual AI, which can return value toward the development of ethical AI and assuring equitable access to AI technologies.