Wildfires are increasing in frequency and severity due to climate change, land-use changes and droughts, threatening global forests. Traditional detection methods lack speed and scalability. Smart forests, driven by machine learning (ML), utilize sensor networks and data analytics to facilitate early warning systems, risk evaluation and mitigation strategies. This article reviews recent ML advancements in wildfire detection, prediction and response, emphasizing current technologies, real-world applications and key challenges, while outlining future opportunities for enhancing forest resilience through intelligent, data-driven solutions.