While others theorize, Dr. Benoît Rizk is hands-on. For the last five years he has been implementing AI at scale, across 20 private imaging centers, with real clinicians, real data, and real consequences when things go wrong.
As Chief Medical Innovation Officer at the 3R Swiss Imaging Network, one of Switzerland's largest private radiology groups, he sits at the intersection where vendor promises meet clinical reality.
Dr. Rizk knows firsthand that models are not the hard part, but governance and integration. Getting AI results to the radiologist at exactly the right moment in the workflow, not a second too late.
In a video interview with AuntMinnie Europe, he addresses what he calls the sycophancy problem in generative AI, the cybersecurity risks that grow as more tools connect to hospital networks, and the real probability that certain roles will be significantly altered within the next decade as automated reporting pipelines mature.
Video/Photo produced by Christof.G.Pelz | GRAFIFANT Creation.Grafik.Photo | www.grafifant.at | 2026.
At SIIM 2026 on June 11, Dr. Rizk will present alongside Dr. Merel Huisman and Dr. Amine Korchi with real-world implementation lessons from a network running more than ten clinical AI solutions simultaneously, alongside a deep dive into EU AI Act compliance, MDR alignment, and EHDS requirements. The session takes place 4:15 till 5:00 p.m. ET, Room 304, Level 3.




![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnieeurope.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&fit=crop&h=100&q=70&w=100)






![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnieeurope.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&fit=crop&h=112&q=70&w=112)







