The AI landscape doesn't move in one direction — it lurches. Some techniques leap from experiment to table stakes in a single quarter; others stall against regulatory walls, technical ceilings, or organisational inertia that no amount of hype can dislodge. Knowing which is which is the hard part. The State of Play cuts through the noise with a rigorously maintained index of AI techniques across every major business domain — classified by maturity, evidenced by real-world adoption, and updated daily so you always know where you stand relative to the field. Stop guessing. Start knowing.
A daily newsletter distilling the past two weeks of movement in a domain or two — delivered to your inbox while the index updates in the background.
Each dot marks the weighted maturity of practices within a domain — hover for a brief summary, click for more detail
Surgical robots that perform portions of procedures autonomously under surgeon supervision. Includes autonomous suturing and tissue manipulation; distinct from surgeon-assisted systems which augment rather than autonomously perform. Scope covers ML/AI-driven approaches; prior deterministic or rules-based automation is out of scope.
Semi-autonomous surgical robotics has crossed from laboratory demonstration into early commercial deployment, but production use remains rare and concentrated at specialist centres. Unlike the widely installed surgeon-assisted platforms that augment human control, this practice involves robots performing discrete procedural steps -- suturing, tissue manipulation, retraction -- independently under surgeon supervision, driven by ML/AI rather than deterministic automation. The technical case is compelling: autonomous systems have matched or exceeded human surgeons on ex vivo subtasks, and at least one dedicated semi-autonomous platform now operates commercially across thousands of patients. Yet the gap between technical feasibility and clinical proof is the defining tension. No randomised trial has demonstrated that autonomous execution improves patient outcomes, regulatory pathways for autonomous surgical features remain undefined, and recurring platform reliability issues reinforce caution. The field sits squarely at bleeding-edge: real deployments exist, credible results are accumulating, but widespread adoption carries unresolved safety, liability, and evidentiary risks.
Moon Surgical's Maestro is the clearest signal that semi-autonomous surgery has moved beyond the lab. The dedicated laparoscopic platform has reached 2,200+ patient deployments worldwide across 60 surgical types, with a multi-centre IDEAL 2a study reporting zero device-related adverse events across 45 solo abdominal procedures. Institutional adoption is accelerating: CHU de Rouen became the first French university hospital to deploy the system in February 2026, and adoption is expanding across Europe and into the United States following FDA clearance. Seger Surgical's first-in-human trials of semi-autonomous intracorporeal anastomosis (March 2026) completed 3 successful bowel closures at 100% success rate in under 2 minutes, signaling ecosystem expansion beyond Moon Surgical with additional vendors entering the semi-autonomous surgical space. Meanwhile, Intuitive Surgical's da Vinci 5 -- with 1,200 systems installed and 270,000+ procedures performed -- provides the computing infrastructure onto which autonomous features could eventually layer, though the platform itself remains surgeon-controlled.
Research continues to push boundaries. Johns Hopkins and Stanford's SRT-H system achieved 100% accuracy on autonomous gallbladder removal in ex vivo models, and UCSD's image-to-grasp pipeline demonstrated state-of-the-art suture manipulation across 400+ trials. Novel ML frameworks are advancing task execution: SutureAgent (March 2026) achieved 58.6% error reduction in autonomous needle trajectory prediction on patient-derived datasets, bridging the gap from simulation to clinical-grade ML. Large-scale deployment evidence demonstrates adoption momentum: China documented 12,000+ robotic surgeries and 800 remote 5G-enabled surgeries, signaling mainstream infrastructure readiness.
The obstacles are as conspicuous as the progress. Intuitive cancelled a software integration partnership with Surgical Science in February, reverting to selective deployment -- a signal that ecosystem maturation is slower than anticipated. Platform reliability remains a critical barrier: the Amy J. Reed Collaborative petitioned the FDA for a Class I recall of da Vinci 5 over intraoperative power failures, and a separate Class II recall followed weeks later for console software defects. An FDA early alert (March 2026) documented da Vinci SureForm 30 stapler malfunction resulting in 4 serious injuries and 1 death, reinforcing equipment safety as a live concern in production deployment. Critical research synthesis reveals regulatory gaps: analysis found 14-fold increase in adverse events after AI integration in surgical guidance systems and fewer than 2% of 1,357 FDA-authorized AI-enabled devices had supporting randomized trials. No randomised trial has yet shown that autonomous execution improves patient outcomes, and liability frameworks for surgeon-supervised autonomy remain undefined. Real-world adoption barriers include automation bias, where clinicians over-defer to system recommendations, and patient/family uncertainty about trusting machine-directed care. Robotic surgery itself accounts for only 5% of US procedures and 2% in Europe, underscoring how narrow the runway is for semi-autonomous features that depend on that installed base.
— King's College London experts forecast surgical robotics shift toward autonomous systems with human-in-the-loop control, embodied AI, and autonomous surgical assistance—establishing the technical trajectory for semi-autonomous procedure adoption.
— Johns Hopkins expert H. Russell Taylor frames core challenges delimiting feasible semi-autonomous surgery scope: specifying clinical intent to autonomous systems and ensuring safe execution within defined constraints; proposes hybrid approaches combining AI with patient-specific digital twins.
— Irish HPRA (regulatory authority) issued device modification guidance for da Vinci X & Xi systems in April 2026, documenting ongoing operational/safety concerns affecting deployed installed base and constraining autonomous feature integration.
— Q1 2026: da Vinci procedures grew 16% to 847K cases, Ion surged 39% to 43K cases; management raised 2026 guidance to 14-16% da Vinci growth, though China policy uncertainty and GLP-1 drug competition signal market headwinds.
— Analyst intelligence documents multiple Class II FDA recalls of da Vinci reusable instruments (broken pitch cables across monopolar scissors, graspers, drivers, hooks, forceps) with worldwide distribution and ongoing status, signaling systemic quality concerns in deployed base.
— Peer-reviewed analysis of autonomy classification systems and governance frameworks for autonomous surgical robots; proposes mechanisms for responsibility attribution, algorithm transparency, and safety monitoring as prerequisites for responsible innovation.
— FDA clearance of Moon Surgical's ScoPilot—AI-driven autonomous camera positioning on Maestro platform. First FDA-cleared AI intraoperative function running natively on commercial surgical robotics, deployed across 1,100+ patients.
— Market analysis projecting microsurgery robot market at 17.12% CAGR for semi-autonomous systems (fastest-growing segment) through 2031. Documents 900+ clinical cases in Europe; semi-autonomous systems reducing proficiency curve from 40 manual cases to 15 cases.
2020: First clinical trial data on autonomous robotic-assisted urology surgery published; preclinical research validated autonomous micro-suturing (0.27mm accuracy) and semi-autonomous tissue retraction with deep learning; 83% of surgeons expressed willingness to adopt AI-based autonomous surgical assistants; emerging concerns about system reliability and complication rates in production robotic platforms.
2021: First in vivo autonomous laparoscopic surgery (intestinal anastomosis) demonstrated on porcine models with level 3 autonomy, outperforming human surgeons on consistency and accuracy; imitation learning approaches enabled autonomous surgical tasks (tissue manipulation, needle handling, knot-tying) on da Vinci; human clinical trials on autonomous cardiac interventions registered. Concurrently, research community assessment reveals critical maturity gaps: systematic review of 35 AI studies found no evidence that AI can identify patient outcome-determining surgical tasks, and clinical researchers increasingly advocate for enhancement-focused collaboration over autonomous execution.
2022-H1: Johns Hopkins STAR robot achieved first autonomous laparoscopic intestinal anastomosis on pig tissues with results superior to human surgeons, demonstrating significant technical progress. Field expanded beyond task-specific research: systematic reviews confirmed supervised autonomous systems are active research area facing ethico-legal hurdles; neurotechnology-enabled semi-autonomous assistance (EEG-driven suction support) showed promise for surgeon-robot collaboration. However, critical adoption barriers emerged: FDA MAUDE database analysis revealed 84.5% instrument malfunction rates and 15.6% patient injury rates in existing da Vinci platforms; legal malpractice analysis identified substantial liability risks that inform autonomous system regulatory trajectory. Industry standardization efforts began with expert consensus on AI and autonomy definitions, signaling field maturation.
2022-H2: Research momentum continued with robotic systems demonstrating improved microsurgical precision in wristed instruments and autonomous suturing frameworks winning academic competitions. Johns Hopkins and institutional research mapped autonomy trends across 4,000+ telerobots in daily use. However, adoption barriers solidified: international survey of 2,191 respondents revealed persistent liability and blame-allocation concerns even in fully autonomous scenarios; systematic review identified deep learning barriers including computational demands, data scarcity, and surgeon knowledge gaps. Field remains at research-to-clinical transition stage with mature technical capability but unresolved governance and regulatory pathways.
2023-H1: Moon Surgical's Maestro semi-autonomous laparoscopic system achieved CE Mark regulatory approval (April 2023) with a 50-patient clinical trial underway in Brussels, marking the first regulatory approval of a dedicated semi-autonomous surgical assistance platform. Concurrent platform reliability concerns emerged: German health authorities issued urgent field safety notices for da Vinci systems (March 2023), indicating ongoing device malfunction and corrective action requirements that constrain autonomous system integration. Research ecosystem continued to expand but adoption barriers persisted: surgeon interest in AI-based autonomous assistance remained high, yet no peer-reviewed evidence demonstrated outcome benefits, regulatory pathways for autonomous systems remained undefined, and liability frameworks remained unsettled.
2023-H2: Research acceleration demonstrated tangible progress in autonomous subtasks: Johns Hopkins published advances in single-arm autonomous suturing with integrated thread management and autonomous vaginal cuff closure outperforming human surgeons in synthetic tissue validation; independent research validated mental workload-based semi-autonomous suction assistance and soft tissue retraction via reinforcement learning. Intuitive Surgical's 2023 results showed sustained da Vinci platform adoption (14 million cumulative procedures, 22% annual growth), indicating continued market expansion. However, adoption barriers remained entrenched: no randomized trials proved semi-autonomous execution improves patient outcomes; regulatory pathways for autonomous features stayed undefined; platform reliability issues persisted; and liability frameworks lacked clarity, constraining deployment beyond clinical trials and research settings.
2024-Q1: Major commercialisation milestone: Intuitive Surgical released da Vinci 5 (FDA March 2024) with 10,000x greater AI-enabling computing power and force feedback reducing tissue trauma by 43%, positioning the platform for autonomous feature integration at scale. Moon Surgical's Maestro reached commercial deployment with 200+ patients treated across 2 French hospitals, demonstrating transition from clinical trial to multi-site adoption of dedicated semi-autonomous laparoscopic assistance. Regulatory momentum continued. However, fatal injury lawsuits and ongoing device recalls reinforced safety and liability concerns constraining adoption; clinical benefit remained unproven with no completed randomized trials.
2024-Q2: Systematic reviews documented regulatory and technical maturity in the field: FDA autonomy classification review found regulatory frameworks lagging behind technical capability (86% FDA-cleared systems at Level 1 autonomy, only 2 with ML capabilities), while autonomous suturing literature showed sustained 41-paper research ecosystem demonstrating feasibility progress in core subtasks. Persistent reliability barriers remained: legal analysis identified mechanical failures, software glitches, and programming errors as common complications constraining adoption. Regulatory approval frameworks for autonomous features and clinical outcome evidence remained the primary barriers to scale beyond specialist centres.
2024-Q3: Research and commercial momentum accelerated around autonomous surgical task execution. Major institutional funding commitment: ARPA-H awarded up to $12M to a multi-institution consortium (Vanderbilt, Johns Hopkins, Utah) for developing fully autonomous surgical robot for tumor removal, with 3-year demonstration goal in simulated conditions. Autonomous suturing research reached state-of-the-art: UC San Diego's image-to-grasp pipeline demonstrated over 400 grasping trials with high accuracy for autonomous thread manipulation. Autonomous needle steering for MRI-guided prostate procedures advanced feasibility in critical surgical applications. ICRA 2024 hosted a full-day workshop on autonomy in surgical robotics, bringing together industry and academic leaders to address regulatory challenges (EU AI Act) and clinical integration pathways. Commercial deployments evolved: Moon Surgical's Maestro reported 250+ patients treated across multiple hospitals (up from 200 in Q1), and Medical Microinstruments' Symani received FDA De Novo clearance for reconstructive microsurgery with first U.S. clinical cases completed. Intuitive Surgical's Q2 patent filings showed continued focus on autonomous procedure stage determination for context-aware surgical assistance. Regulatory frameworks and unproven clinical benefit remained principal barriers to scaling beyond specialist centres and research-active institutions.
2024-Q4: Field matured toward clinical deployment and industry standardization. Johns Hopkins demonstrated major imitation learning breakthrough: robots trained directly from surgical video achieved human-equivalent performance without manual programming, reducing dependency on hand-coded procedures and advancing feasibility of generalizable autonomous surgical systems. Moon Surgical's Maestro completed first-in-human clinical validation with 30 successful laparoscopic procedures at CHU Saint-Pierre teaching hospital, transitioning from research trial to clinical evidence. Regulatory landscape clarified through market analysis documenting divergence between FDA's SaMD approval pathway and EU Medical Device Regulation (MDR) requirements, identifying harmonization challenges constraining international deployment of AI-enabled autonomous systems. Expert consensus from Johns Hopkins robotics panel emphasized human-machine complementarity and surgeon oversight as realistic near-term pathway rather than full autonomy, reflecting tempered expectations within leading research institutions. Platform reliability concerns persisted, constraining adoption momentum despite technical progress.
2025-Q1: Vision-language imitation learning advanced autonomous task execution: researchers integrated vision-language models trained on 20 hours of surgical video with da Vinci robots, achieving zero-shot autonomous performance on tissue manipulation, needle use, and suturing with problem-solving capability (autonomous needle retrieval). Reinforcement learning frameworks for autonomous suturing demonstrated higher accuracy than teleoperation on synthetic tissue. Intuitive Surgical's da Vinci 5 achieved broad market adoption (2,500+ surgeons, 493 units in Q4 2024, 40+ procedure types), establishing the computing platform foundation for autonomous feature integration. However, platform reliability challenges emerged: Class 2 FDA recall (March 2025) of da Vinci 5 due to faulty foot pedal springs (439 units affected) reinforced safety and liability barriers constraining autonomous deployment despite technical progress. Field remained at clinical validation stage with mature autonomous subtask demonstrations but unresolved platform reliability and outcome evidence gaps.
2025-Q2: Moon Surgical's Maestro semi-autonomous laparoscopic system reached inflection-point commercial adoption with 1,600+ total procedures (1,400 in preceding year) across 60 surgical types, demonstrating rapid scaling from clinical trials into multi-site ambulatory deployment. Second-generation system rolled out in France with advanced AI for setup automation and vision control, and company expanded leadership teams for U.S. commercialization. Institutional research validated semi-autonomous execution across surgical domains: translational study demonstrated superior accuracy in dental implant placement with zero complications across 60 human participants. Industry projections predicted global autonomous surgical robotics market exceeding €10 billion by 2033, reflecting sector confidence. However, regulatory pathways for autonomous features remained undefined, no completed randomized trials demonstrated clinical benefit, platform reliability concerns persisted, and liability frameworks remained unresolved—constraining adoption beyond specialist centres and research institutions.
2025-Q3: Major breakthrough in autonomous surgical procedure comprehension: Johns Hopkins and Stanford's SRT-H AI system achieved 100% accuracy performing autonomous gallbladder removal on ex vivo models, advancing from task-specific execution to procedural understanding. Concurrent academic progress in autonomous suturing: UCSD pipeline continued refining suture manipulation with 400+ trials demonstrating state-of-the-art accuracy, and NeurIPS 2025 introduced SutureBot benchmark (1,890 suturing demonstrations) enabling 59-74% precision improvements. Systematic review of 32 studies documented growing research momentum in autonomous surgical assistants, with emerging progress in tool manipulation and persistent challenges in procedural awareness. Moon Surgical's Maestro continued commercial scaling with 1,600+ global procedures across 60 surgical types. Critical barriers remained: regulatory frameworks undefined, clinical outcome evidence absent, platform reliability concerns ongoing, and liability standards unresolved—limiting deployment to specialist centres and research institutions.
2025-Q4: Mainstream platform adoption accelerated with da Vinci 5 placements doubling (240 units in Q3 vs. 110 in Q3 2024); Huntsman Cancer Institute deployed full 10-system fleet with reported 40% tissue force reduction. Ecosystem infrastructure matured through integration of simulation training software into all da Vinci 5 systems. Research continued advancing autonomous capabilities: SRT-H gallbladder autonomy maintained 100% accuracy; UCSD suturing pipeline demonstrated 400+ trials of state-of-the-art performance. Critical expert commentary emphasised gap between technical breakthroughs and clinical readiness, highlighting human factors and biological variability unaddressed in ex vivo validation. Platform reliability concerns persisted: Class 2 FDA recall issued December 2025 (144 da Vinci 5 units) for Surgeon Console connector. Regulatory expansion occurred with FDA clearance of da Vinci SP for three general surgery procedures. Critical adoption barriers remained entrenched: no randomized trials proving clinical benefit, no regulatory pathways for autonomous feature integration, liability frameworks undefined, constraining deployment beyond specialist centres despite mature technical capability and emerging market traction.
2026-Jan: Mainstream platform regulation expanded with FDA 510(k) clearance for da Vinci 5 cardiac procedures, positioning platform for broader autonomous feature integration. Moon Surgical's Maestro achieved 2,200+ patient deployments worldwide with FDA-approved iterative AI updates, demonstrating commercial viability of dedicated semi-autonomous systems. IDEAL 2a multi-center study validated safety and feasibility of Maestro for solo abdominal surgery across 45 procedures with zero device-related adverse events, advancing clinical evidence. Platform adoption momentum continued with 1,200 da Vinci 5 systems installed (270,000 procedures globally), yet critical adoption barriers remained largely unchanged: no clinical benefit evidence, undefined liability frameworks, and ongoing platform reliability concerns constraining deployment beyond specialist centres.
2026-Feb: Institutional adoption of dedicated semi-autonomous platforms accelerated: CHU de Rouen (France) became first university hospital to adopt Moon Surgical's Maestro, with first procedure February 3, 2026, demonstrating expansion across major teaching institutions. Global adoption metrics confirmed: ACS bulletin documented 5% robotic surgery penetration in US, 2% in Europe, underpinning infrastructure for semi-autonomous feature integration. Ecosystem challenges emerged: Intuitive Surgical cancelled software integration MOU with Surgical Science on February 2, reverting to selective deployment to minority of da Vinci 5 installed base, signaling adoption velocity constraints. Critical platform safety barriers reinforced: FDA petition from Northeastern University's Amy J. Reed Collaborative (February 9) requested Class I recall of da Vinci 5 due to intraoperative power failures; Class II FDA recall issued February 25 for Surgeon Console software configuration defects. Regulatory expansion and platform reliability concerns provided simultaneous signals of market growth and unresolved safety constraints.
2026-Mar/Apr: Institutional mainstream recognition elevated semi-autonomous surgery to national-level discourse. National Academy of Sciences Distinctive Voices lecture (April 1) explicitly stated 'autonomous surgery is no longer science fiction' and documented 'semi-autonomous robots perform surgery' with 'transformative potential.' Novel ML frameworks advanced autonomous suturing: SutureAgent (arXiv March 2026) reduced needle trajectory prediction error 58.6% on patient-derived datasets (50 patients, 1,158 trajectories), demonstrating technical progress beyond simulation. Commercial ecosystem expanded: Seger Surgical completed first-in-human semi-autonomous bowel closure (100% success, <2 min closure vs. 15-20 min manual), signaling second company entering semi-autonomous surgical deployment. Moon Surgical's ScoPilot achieved FDA clearance for autonomous camera positioning on Maestro — the first FDA-cleared AI intraoperative function running natively on a commercial surgical robot platform, with 1,100+ patients treated; Medtronic deployed Stealth AXiS with real-time autonomous vertebral tracking for complex spine fusion, extending autonomous navigation beyond laparoscopy. Competing ecosystem matured: Chinese systems (MP1000/MP2000, Edge Cloud) received NMPA regulatory clearance with reported 100% multi-specialty success rates; microsurgery robot market analysis projects 17.12% CAGR through 2031, with semi-autonomous systems as the fastest-growing segment reducing surgeon proficiency curves from 40 to 15 cases. International consensus standards for robotic stroke treatment (King's College London) assessed AI-assisted mechanical thrombectomy robotics as 'not yet ready for routine clinical use' despite hardware-level autonomy demonstrations — a meaningful check on adoption pace in vascular applications. Critical safety concerns persisted: FDA documented da Vinci curved-tip stapler failures (4 injuries, 1 death) and UK MHRA issued field safety notice for da Vinci X & Xi systems, signaling ongoing reliability constraints on incumbent platforms. Research synthesis identified regulatory gaps: critical analysis found 14-fold increase in adverse events after AI integration and <2% of 1,357 FDA-authorized AI-enabled devices supported by randomized trials, exposing barriers to autonomous surgical system maturation. Field remains at inflection point: multiple commercial platforms now deployed, institutional recognition mainstream, but platform safety and undefined regulatory frameworks continue constraining rapid scaling.
2026-May: Technical trajectory and governance frameworks crystallized. King's College London perspective (May 7) forecasted surgical robotics shift toward autonomous systems with human-in-the-loop control and embodied AI, establishing the forward technical vision for semi-autonomous procedure integration. Expert consensus solidified: Johns Hopkins authority H. Russell Taylor framed core challenges delimiting feasible autonomous scope—specifying clinical intent and ensuring safe execution within constraints; proposes hybrid approaches combining AI with patient-specific digital twins. Institutional innovation analysis from Beijing University assessed autonomy classification systems and governance frameworks; proposes responsibility attribution, algorithm transparency, and safety monitoring mechanisms as prerequisites for safe autonomous deployment. Platform adoption metrics remained strong: Intuitive Surgical Q1 2026 showed 847K da Vinci procedures (16% growth) and 43K Ion procedures (39% growth), with 2026 guidance raised to 14-16% da Vinci growth; however, China policy uncertainty and GLP-1 drug competition signal structural headwinds. Critical platform safety barriers persisted: Irish HPRA (May 5) issued device modification guidance for da Vinci X & Xi systems; analyst intelligence documented multiple Class II FDA recalls of da Vinci reusable instruments (broken pitch cables across monopolar scissors, graspers, drivers, hooks, forceps) with worldwide distribution and ongoing status, signaling systemic quality concerns in deployed base constraining autonomous feature integration momentum. Governance and frameworks advanced; platform safety and clinical evidence gaps remained principal barriers to rapid scaling beyond specialist centres.