This paper addresses the “electrification paradox”: while electric vehicles (EVs) are vital for decarbonising transport, the rapidly growing wave of end-of-life (EoL) EV batteries creates a new, high-stakes hazardous waste stream that threatens the resilience of storage, transport and recycling infrastructures. We first synthesise the regulatory and scientific basis for treating EoL lithium-ion batteries as hazardous materials, highlighting the combined risks of toxic emissions, long-term environmental contamination, and acute thermal runaway events. Building on this risk profile, we propose an integrated technology framework that combines the Internet of Things (IoT), high-fidelity Digital Twins, and Generative Artificial Intelligence (AI) to move from reactive, compliance-driven management to predictive, adaptive and resilient operations. Three detailed use cases—proactive thermal-runaway prevention in warehouses, dynamic rerouting of high-risk loads, and AI-assisted emergency response—illustrate how the framework senses, simulates, and responds to evolving hazards in real time. We then examine key implementation challenges, including cybersecurity, scalability, interoperability, explainability, human-in-the-loop governance and algorithmic fairness. Finally, we show how the same data backbone underpins a Digital Battery Passport that enables second-life markets, optimised recycling and alignment with emerging policy, positioning EoL batteries as traceable resources within a secure circular economy. We conclude by outlining future research directions in multi-agent control, quantum optimisation and end-to-end digital threads for critical materials.