The integration of drones into Internet of Things (IoT) ecosystems has revolutionized various domains, offering unparalleled mobility and adaptability for data collection, monitoring, and surveillance. However, drones are susceptible to security vulnerabilities, stemming from network infrastructure weaknesses, software vulnerabilities, and potential malicious intent. Unsecured drones are vulnerable to hacking, data theft, and manipulation, posing risks to public safety and privacy. Protecting against these susceptibilities is paramount to ensure the safe and responsible integration of drones into society. Incorporating sophisticated Machine Learning (ML) techniques into drone networks presents a promising solution to mitigate security risks effectively. ML algorithms analyse real-time data to detect anomalies and predict threats, enabling prompt responses to potential risks. Machine Learning powered drones can autonomously adapt security measures to evolving threats, enhancing encryption protocols to resist tampering and interception. The chapter explores the recent privacy and security concerns surrounding drone networks and discusses the integration of ML-based security solutions to alleviate risks. Challenges and opportunities in ML-driven drone security are also explored, emphasizing the importance of ensuring safe and secure drone operations in the evolving landscape of IoT ecosystems.