中国医疗健康：AI 2.0

Equity ResearchJuly 8, 2026 China (PRC) | Healthcare AI 2.0: Healthcare's Next Prescription Why are MNCs writing checks to almost every Chinese AIDD company?It's not about certainty, it's about exposure - MNCs are investing becausetheycan't afford not to.HC AI is moving from experimentation tocommercialization,where we see four themes emerging:AIDD,BCI,intelligent medtech, and HC AI agents; of which we believe AI-native biotechplatforms offer the most compelling risk-reward.Buy Insilico, METiS,Mindray, UIH, MedBot, JDH. In our AI Healthcare 1.0 report, we highlighted AI's potential to improve accessibility, accuracy,and efficiency across healthcare. The industry has since moved beyond experimentation, withmilestones such as Insilico's Rentosertib advancing into Phase 3 and China's first approvedBCI product for paralysis. As AI enters clinical validation and commercialization, we see fourinvestable themes emerging: AIDD, BCI, intelligent medtech, and healthcare AI agents. AIDrug Discovery=Data+Algorithms+Workflow,Driving Faster,Smarter,and MorePredictable Drug R&D-Healthcare AI is more than a science project - it's becoming acommercial reality. In our view, the biggest risk in healthcare AI may no longer be investing,but not investing. WithInsilicoadvancing into Ph3 and China's first BCI approved, AI is movingfrom concept to commercialization. AIDD is evolving from software-driven efficiency to anAI + biotech asset creation story. Byintegrating data, algorithms, and automated workflows into a closed-loop R&D engine, AI couldreduce drug discovery timelines by70-90%and save>US$50bnof value across the drugdevelopment chain. Our survey indicated AI adoption is broadening across the pharma valuechain, with the highest penetration in discovery (67-70%) and manufacturing (82%), followedby clinical data mgmt (56-67%) and regulatory intelligence (44%). We believe the next generation of winners will be defined not by model performance alone, butby their ability to translate AI-generated insights into differentiated drug assets viaproprietarydata,closed-loop dry-wet labs, andin-house development capabilities. Drug discovery'siterative design-test-learn cycle is a natural fit for AI, and China offers a compelling environmentfor this model, supported by deep scientific talent, rich data generation, and cost-efficientexperimentation. We prefer theAI + biotechmodel, as it captures value from both platformmonetization and proprietary pipelines.Related cos: Insilico, METiS, Xtalpi. Cui Cui, CFA * | Equity Analyst+852 3767 1228 | cui.cui@jefferies.com David Shang * | Equity Analyst+852 3743 8017 | david.shang@jefferies.com David Windley, CFA ^ | Equity Analyst+1 (615) 963-8313 | dwindley@jefferies.com Brain-Computer Interfaces: Decoding Thoughts into Therapies-BCIs are progressing fromrehabilitation devices to therapeutic platforms. While still early, advances in AI and neuraldecoding are bringing clinical adoption closer, with invasive and ultrasound-based interfacesrepresenting the most promising innovation frontiers.Related cos: BrainCo, Neuracle, StairMed. Tycho Peterson ^ | Equity Analyst+1 (212) 738-5583 | tpeterson2@jefferies.com Edison Lee, CFA * | Equity Analyst852 3743 8009 | edison.lee@jefferies.com Medtech: Turning Hardware into Intelligent Platforms-AI is reshaping medtech acrossimaging, robotics, and diagnostics, moving the industry from hardware-centric products tointelligent platforms. We believe leaders such as UIH, Mindray and MedBot are best positionedto benefit from this transition through integrated "AI + equipment" solutions that improveclinical outcomes and operating efficiency.Related cos: Mindray, United Imaging, MedBot. Matthew Taylor, CFA ^ | Equity Analyst1 (212) 778-8721 | matt.taylor@jefferies.com Julien Dormois ‡ | Equity Analyst+33 1 8665 6375 | jdormois@jefferies.com James Vane-Tempest § | Equity Analyst44 (0) 20 7029 8275 | jvane-tempest@jefferies.com AI Agents: From Chatbot to Healthcare Operating System-Healthcare AI agents are movingup the value chain - from consumer chatbots to physician copilots and ultimately hospitaland payer infrastructure. Each step unlocks larger budgets, deeper workflow integration,and stronger monetization opportunities, shifting AI from a traffic-generation tool to a coreoperating layer of the healthcare system.Related cos:JD Health, Ali Health, PAGD. Bingyu Chen * | Equity Analyst+852 3767 1346 | bingyu.chen@jefferies.com Hongda Zhong * | Equity Analyst+852 3767 1311 | hongda.zhong@jefferies.com Jing Ma, Ph.D. * | Equity Analyst+852 3767 1277 | jing.ma@jefferies.com Table of Contents Part I: AI drug R&D – Disruption or Breakthrough? 1.1 AIDD=Data+Algorithms+Workflow The core of AIDD is the transformation of traditional, fragmented drug development into a data-driven, closed-loop system that replaces "random discovery" with "rational design." By integratingthree most important elements, big data, algorithms, and workflow, AIDD seeks to so