Tobacco use remains one of the leading preventable causes of death, with traditional smoking-cessation treatments having low success rates for sustained abstinence (<10%), which highlights the need for innovative strategies. This paper examines opportunities to combine AI and machine learning technologies with interventions designed to help people quit smoking, particularly the potential for real-time data from wearables, covering behavior, physiology and other personal factors, to improve evidence-based cessation approaches within US healthcare systems. The study used systematic review methods that include RCTs, real-world evidence from health systems, regulatory issues and demographics. The paper summarizes current evidence on AI-powered digital health interventions for smoking cessation. The findings indicated that though conversational AI chatbots and machine learning recommender systems show early signs of effectiveness, significant implementation challenges remain, such as HIPAA regulation complexities, disparities caused by fragmented mHealth data standards, limited evidence on cost-effectiveness and wide variation in wearable device adoption across different age, income, education and racial/ethnic groups. Major US surveillance systems like BRFSS and PATH, along with integrated healthcare organizations such as Veterans Affairs and Kaiser Permanente, demonstrate that systematic methods substantially increase evidence-based treatment delivery, but AI applications in this area are still in early development. The review concludes that realizing AI's full potential in tobacco cessation requires addressing technical, clinical, regulatory and equity challenges through implementation science approaches focused on real-world effectiveness, sustainability and equitable access for all populations affected by tobacco use.