Samsung ups the AI health game with galaxy watch 9 duo

Samsung’s next-generation Galaxy Watch, set to debut at the July 22, 2026 Unpacked event, represents a bold hardware overhaul designed to support advanced AI-driven health tracking and significantly enhanced battery life. Featuring “all-new internal components,” the Galaxy Watch 9 and Galaxy Watch Ultra 2 will offer improvements such as titanium construction, water resistance, Bluetooth 6.0, and Wear OS 7 with One UI Watch 9, underscoring Samsung’s commitment to premium build quality and durability while enabling continuous, real-time wellness monitoring.

At the core of Samsung’s strategy is positioning the Galaxy Watch as a 'health-first, always-on gateway' within its broader Galaxy ecosystem, leveraging Galaxy AI to deliver personalized, continuous background monitoring and real-time, AI-powered health insights. Moving beyond traditional step tracking, the watch aims to transform raw sensor data into proactive, actionable intelligence that helps users manage their health before issues escalate, reflecting a deeper AI integration across Samsung’s wearable lineup and echoing cross-device AI ambitions seen in the industry.

A critical upgrade in this new generation is the enhanced battery life, addressing a longstanding limitation of feature-rich smartwatches. Samsung’s focus on improved endurance is designed to support the watch’s always-on health monitoring capabilities without compromising all-day wearability, positioning it as a serious contender against the Apple Watch’s battery performance and reinforcing the device’s role as a reliable, intelligent health companion.

At Galaxy Unpacked 2026, Samsung unveiled a visionary Connected Care ecosystem that aims to transform healthcare from reactive treatment to proactive, personalized prevention by leveraging continuous monitoring and AI-driven insights. Dr. Hon Pak highlighted the importance of bridging the 'first mile' of early detection outside clinical settings with the 'last mile' of daily individual support, emphasizing that Samsung’s approach goes beyond raw data to understand contextual moments when users are ready to change or need encouragement. This shift is designed to close the gap between knowing and doing by making healthy choices more accessible through AI-powered personalization, as exemplified by Samsung Health’s integration of meal logs, the Antioxidant Index, and a global recipe library to foster behavior change rather than mere tracking, a strategy detailed by Kartik Poria.

Samsung’s Connected Care strategy is structured around three pillars—everyday behavior, clinical credibility and privacy, and advanced AI and research—underscoring the critical role of partnerships in building a scalable preventive health ecosystem. A prime example is the collaboration with iFIT, whose CEO Kevin Duffy described how expert-led fitness programs integrate seamlessly with Galaxy Watch and Samsung Health, delivering a holistic Fitness Index and personalized recommendations that support both physical and mental wellbeing. This ecosystem approach not only enhances accessibility but also enriches the personalized care experience, aligning with Samsung’s broader vision of connected, continuous, and credible health management.

Samsung Research America has pioneered two foundation AI models, xMAE and HiMAE, that revolutionize wearable health monitoring by enabling real-time biosignal analysis directly on smartwatches without cloud dependency. HiMAE impressively processes health data in under one millisecond on smartwatch CPUs by analyzing biosignals across multiple time scales simultaneously, while xMAE reconstructs ECG data from continuous PPG signals to deliver precise cardiovascular insights. This breakthrough allows for continuous, privacy-focused health monitoring with minimal battery impact, addressing critical challenges in latency, privacy, and energy efficiency.

Both xMAE and HiMAE leverage self-supervised learning on vast amounts of unlabeled biosignal data, a technique borrowed from the language model revolution, enabling them to generalize across diverse health tasks such as cardiovascular disease prediction and sleep classification. Notably, xMAE was pretrained on approximately 9,400 hours of paired ECG and PPG data, outperforming existing single-signal and multimodal models in 15 out of 19 evaluation tasks, which underscores its ability to bridge the accuracy gap between ECG and PPG sensors for passive, continuous cardiovascular monitoring without active user input.

Samsung’s shift from cloud-based to edge AI health monitoring marks a fundamental transformation in wearable technology, tackling the triad of privacy, latency, and battery life concerns that have long hindered real-time biosignal analysis. By processing sensitive health data entirely on-device, Samsung eliminates the need for constant server communication, thereby preserving user privacy and extending battery life while enabling instantaneous health interventions. The acceptance of xMAE and HiMAE research at prestigious AI conferences ICML and ICLR further validates the scientific rigor and innovation behind this approach, signaling a new era of predictive, personalized health monitoring powered by advanced on-device AI.

Samsung’s strategic advantage lies in its end-to-end control over both hardware and AI, exemplified by its xMAE and HiMAE models that perform complex health data processing entirely on-device. This approach eliminates cloud dependency, enhancing user privacy and enabling sub-millisecond responsiveness on smartwatch-class CPUs, a feat that sets Samsung apart in real-time wearable health monitoring. However, this technological edge comes with significant challenges, including the rigorous validation of model accuracy across diverse health tasks, efficient battery management to sustain continuous monitoring, and stringent data privacy safeguards—all critical factors as Samsung vies against industry giants Apple and Google.

Samsung’s foundation models, xMAE and HiMAE, push the boundaries of wearable sensor accuracy by reconstructing ECG data from PPG signals and analyzing biosignals across multiple time scales within a single pretrained architecture. By outperforming existing models in 15 out of 19 benchmark tasks, these AI innovations provide a versatile platform capable of supporting a broad spectrum of health monitoring functions. Yet, despite this competitive positioning, Samsung must navigate complex regulatory landscapes and ensure these models generalize effectively in real-world scenarios to maintain trust and meet compliance standards, challenges that remain pivotal in the race against Apple’s and Google’s AI-driven health ecosystems.