Identifying the critical view of safety (CVS) and other safe anatomic landmarks during laparoscopic cholecystectomy (LC) is the cornerstone for avoiding bile duct injuries (BDI). Artificial intelligence (AI), which has infiltrated in the operating room, appears as a promising tool, enabling surgeons to safely dissect during LC. The aim of this study is to investigate the AI models and their performance for identifying these critical structures. A systematic literature review of the PubMed and Google Scholar databases was conducted using medical subject headings (MeSH). Studies presenting the application of AI models for identifying CVS and anatomic landmarks were included and analyzed in terms of performance and reliability. Clinical feasibility trials with preliminary data were separately analyzed. Seventeen studies were found eligible and analyzed for various parameters. Generating AI models for identifying CVS and anatomic landmarks during LC is feasible, while their performance in terms of accuracy, precision and recall has remarkably improved. Regarding their reliability, intersection over union (IoU) and F1/Dice scores have been improved, as well. AI models can be successfully deployed in the operating room, and could assist surgeons in decision-making. Implementation of AI during LC for identifying CVS and important anatomic landmarks appears as a feasible and promising option. Preliminary data are encouraging in terms of performance but still major obstacles and barriers need to be overcome. Whether this will lead to reduced BDIs and enhanced patient safety, requires more well-designed studies. PROSPERO database registration: (UIN: CRD42024557432).