In the era of personalized medicine, an ever increasing emphasis is being placed on prognostication and prediction of surgical outcomes and treatment responses in colorectal cancer (CRC).1 Recent advances in information technology have allowed AI integration into clinical practice. Machine learning (ML) is a subfield of AI in which algorithms iteratively ‘learn’ from data to make predictions based on discerned patterns, without being explicitly programmed to do so. Deep learning (DL) is a further subfield of ML, whereby artificial neural networks (ANN) loosely modelled on the human brain structure, are exposed to large or ‘deep’ datasets, deriving predictive capacity from learned patterns in the data.2 Computer vision (CV) refers to the training of computers to read and interpret images, a field closely involved in the last half-decade with DL approaches. Within Radiomics, AI is making an impact in CRC staging.1 DL platforms have demonstrated the ability to distinguish between T2 and T3 rectal cancer on MRI with 90% accuracy,3 and to detect lymph node disease with a higher accuracy than radiologists.4 The combination of histopathological outcomes with MRI and CT imaging shows promise. It allows models to predict a complete pathological response in rectal cancer patients treated with neoadjuvant therapy and the identification of systemic therapy responders with metastatic CRC.1 Colonoscopy is an area in which AI is already being used. AI algorithms designed to aid polyp detection have been shown to increase adenoma detection rates, even amongst experienced endoscopists,5 thus providing a cost effective method of reducing CRC during population screening.6 DL models can both detect and classify polyps with high sensitivity and specificity.7 The development of AI-assisted clinical tools that can provide decision support such as future endoscopic surveillance intervals are ongoing.5 Such tools can also enhance endoscopy training through improved polyp pattern recognition identification.5 Support systems which provide evidence-based clinical decision making tools for CRC are already available. The Watson for Oncology (WFO) program developed by IBM and the Memorial Sloan Kettering Cancer Centre has been trained to follow the U.S. National Comprehensive Cancer Network guidelines.8 Treatment recommendations suggested by the WFO have shown concordance rates of 80%–90% for CRC management compared to a multidisciplinary team (MDT).8, 9 Although not designed to supplant the MDT, AI tools can become an adjunct for clinicians to increase workload time efficiency, and decrease the risk of errors.8 Though still in its infancy, the utilization of AI tools is becoming increasingly relevant within colorectal surgery, through the intraoperative use of CV10 for operative precision and training. Using laparoscopic image data from multiple rectal resections, DL models have been able to recognize the correct phase of the operation, and also act as an intra-operative guide in the identification of complex anatomical structures, such as the total mesorectal excision plane of dissection.10, 11 A recent pilot study has demonstrated the utilization of DL-based software to automate the surgical skill assessment of various steps within a laparoscopic rectal resection, paving its way for future applications in surgical training.12 Despite these exciting developments, many limitations remain. Significant financial investment and creation of regulatory and legal frameworks will be required to integrate AI into existing healthcare systems.13 Furthermore, AI is constrained by the quality and quantity of available data used to train them. As the majority of relevant studies come from only a few geographical locations, AI models are susceptible to inherent biases related to differing patient cohorts.14 AI has the potential to revolutionize the management of CRC. However, there are many challenges to be addressed before AI programs can become embedded into routine care for CRC. Open access publishing facilitated by The University of Melbourne, as part of the Wiley - The University of Melbourne agreement via the Council of Australian University Librarians. Matthew Y. K. Wei: Conceptualization; writing – original draft; writing – review and editing. Junyao Zhang: Writing – review and editing. Reuben Schmidt: Validation; writing – review and editing. Andrew S. Miller: Writing – review and editing. Justin M. C. Yeung: Supervision; validation; writing – review and editing.