Building a future where AI improves health for all means earning the unwavering trust of the patients, clinicians, and researchers we serve.
As a leader in precision health AI, our technological capabilities are inseparable from our ethical responsibility to develop and deploy AI that is useful to patients, clinicians, and researchers while also safeguarding those who use our solutions.
Our approach to AI responsibility is centered on the following six core principles:
Safety, effectiveness, and real-world monitoring
We build responsible AI solutions to improve health. Our unique dual role as a company designing both AI-enabled products and personalized clinical care programs gives us first hand insight into real-world performance. This vertical integration of technology through care informs our comprehensive framework for developing and deploying responsible AI that is safe, effective, and trustworthy, with clinical and scientific rigor applied to data quality and infrastructure. This approach embeds accountability and human oversight from the start, ensuring our technology benefits patients and clinicians.
We anticipate and mitigate risks through adversarial testing and close real-world monitoring to detect performance drift, avoid unintended consequences, and prevent harm. Where relevant, our in-house experts conduct analytical and clinical validations, assessing our AI solutions' ability to produce safe and effective recommendations.
Continuous learning and adaptability
Healthcare is dynamic, and AI must evolve with it while remaining fit-for-purpose. We commit to responsible improvement of our models and systems over time, informed by new data, emerging science, and user feedback. Adaptation is managed under appropriate oversight to ensure safety, effectiveness, and trustworthiness throughout the lifecycle.
AI should work for everyone
We aim to assess representativeness of datasets used for AI development and real-world performance across multiple dimensions of patients and users, including designing for accessibility. We aim to work with clinicians, patients, and communities so our solutions reflect the needs of those often overlooked.
Transparency
Meaningful transparency entails clear disclosure of AI use, communication about an AI solution’s purpose, capabilities, and limitations, and tracking the data used to develop an AI solution.
We are committed to AI solutions that positively impact all the patients, clinicians, and researchers they are designed to serve. We are intentional about the intended use of our models, document the data used to develop and test AI solutions, and systematically evaluate the AI’s effectiveness, limitations, and monitoring of its real-world performance.
Accountability and auditability designed for healthcare
As a health technology company operating in a highly regulated industry and serving customers across the healthcare ecosystem, we understand how critical it is to develop and deploy responsible, safe AI systems and workflows in a way that can be tracked, documented, and easily audited. As such, we embed legal and regulatory standards fit for healthcare into our processes and design our systems for traceability, including maintaining documentation of our AI system’s design, development, and testing.
Privacy and security by design
Stewardship of health data is a sacred trust. We embed privacy, security, and data stewardship into every stage of AI development. This includes compliance with applicable requirements and standards, application of strong technical safeguards, and data use consistent with patient expectations.
Privacy, security, and proactive protection of sensitive data is an intrinsic component of our systems. We adhere to rigorous legal and ethical requirements for data handling, ensuring that user privacy is respected, and that sensitive information is protected throughout the AI lifecycle.
We implement robust technical controls to protect against a wide range of threats. This includes strong access controls and a continuous vulnerability management program. Security isn't just about data privacy — it’s about safeguarding the operational integrity of our systems to ensure they remain reliable and secure, preventing threats that could compromise patient safety. This systematic and persistent focus on security ensures we can proactively detect and respond to threats, maintaining the highest level of trust and protection.