AI’s hidden risk that no one is talking about

ARTIFICIAL intelligence (AI) has captured the imagination of boardrooms, policymakers and investors across Asia-Pacific. Governments are placing AI at the heart of national strategies, while businesses are racing to embed it into everything from customer service operations to market forecasting and supply chain optimisation.

Yet, amid this wave of adoption, there is a growing risk that remains largely overlooked. It is not AI model hallucinations or ethical dilemmas. Instead, it is the quiet accumulation of cyber risk in the “building blocks” of AI development.

The open-source frameworks and libraries that have made AI accessible and scalable are now also one of its greatest vulnerabilities.