Background: Conventional mechanical ventilation (MV) requires frequent manual titration and may expose patients to hypoxaemia, hyperoxia, and injurious settings. Closed-loop and AI-guided systems automate adjustments to support lung-protective ventilation and reduce workload. We aimed to synthesize clinical evidence comparing closed-loop or AI-guided ventilation with conventional strategies. Methods: Following PRISMA, we included randomised or prospective studies of invasively ventilated patients comparing automated (INTELLiVENT‑ASV, SmartCare/PS, automated oxygen control or e‑alert–driven FiO2 titration) vs manual care. Outcomes included quality of ventilation/oxygenation, patient-centred outcomes, safety, and workload. Narrative synthesis was used. Results: Nine original studies were included across adult and paediatric ICUs and post‑operative settings. Automated oxygen control increased time in SpO2 target ranges and reduced time below range without increasing hyperoxia; fully automated ventilation increased time in predefined ‘optimal’ ranges and decreased exposure to injurious settings. Several studies reported fewer manual adjustments and no excess adverse events. Evidence for reductions in duration of ventilation, ICU stay, or reintubation was inconsistent and context‑dependent. Two summary tables present study characteristics and outcomes. Conclusion: Closed‑loop and AI‑guided ventilation are safe and improve control of oxygenation/ventilatory targets, with signals of reduced workload. Larger multicentre trials are warranted to determine effects on patient‑important outcomes.