Our Commitment
AI is integral to Oku's work. We use it to advance understanding of ocular and systemic health, support healthcare professionals and researchers, and contribute to the development of better diagnostics and treatments.
These are the principles of Oku Digital Health, Inc. They apply across Oku and its operating companies, in the development, deployment, and support of our products, our research and data services, and our internal business operations. Each Oku company remains responsible for the regulatory statements, product documentation, and privacy notices that apply to its own products and services.
We are committed to developing and using AI responsibly. These seven principles guide our work across our products, research, services, and business operations.
1. Advance Health and Human Benefit
We apply AI where it can contribute meaningful value to patients, healthcare professionals, researchers, and society. Each use has a defined purpose and considers the people who may benefit from, rely on, or be affected by it.
In practice: we define and document the intended purpose, intended users, and intended population of each system before development proceeds.
2. Maintain Human Accountability
AI can extend human capability, but it does not remove human responsibility. People remain accountable for how AI is designed, developed, deployed, and used. Human oversight reflects the purpose, risk, and level of autonomy of each system.
Some AI systems are designed, validated, and authorized to operate autonomously, producing a clinical result without interpretation by a clinician. Autonomy of that kind does not reduce accountability. It concentrates accountability with us. Where a system is cleared or approved to operate without clinician interpretation, we accept responsibility for its design, its validation, its performance in the field, and the governance and monitoring that surround it, and we make its intended use and its limits clear to the clinicians and organizations that rely on it.
We believe autonomous AI, developed and governed this way, expands access to timely diagnosis in settings where specialist interpretation is not available. Our principles are written to support that work, not to constrain it.
In practice: accountability for each system is assigned to a named function within Oku and recorded in our AI governance documentation.
3. Require Evidence and Safety
We develop and evaluate AI with scientific rigor, appropriate attention to data quality, and a clear understanding of intended use and limitations. Safety, research integrity, and the interests of patients and research participants are central to our work.
The data we use to develop and evaluate AI is traceable to a defined source and to a permitted use. Evaluation does not end at release. We monitor how our systems perform after deployment, investigate issues that arise in real-world use, and update, restrict, or withdraw a system when the evidence requires it.
In practice: intended use, evaluation results, known limitations, and post-deployment monitoring are documented for each system.
4. Promote Fairness and Reliability
We evaluate how AI performs across the populations, data sources, devices, and settings for which it is intended. We work to identify and address avoidable unfair bias and other material performance differences.
Because our systems operate across many care settings, we evaluate performance using the imaging devices, image quality, and workflow conditions actually present where a system will be used, rather than on curated research data alone.
In practice: we define the intended population and operating conditions of each system, and we test against them.
5. Be Transparent
We identify the AI-enabled functions in our own products and services, and we provide information about their purpose, role, and material limitations that is appropriate to the context and intended audience. Where our technology is deployed by a customer, we give that customer the information it needs to meet its own disclosure obligations to clinicians and patients.
In practice: product-specific information about intended use, performance, and limitations is provided in the applicable product labeling and regulatory documentation.
6. Secure Our Systems and Data
We design, build, and operate our AI systems to resist misuse, tampering, and cyber threats, and to remain reliable when conditions change. Security applies across the lifecycle, from the data used in development to the systems our customers depend on in clinical and research settings.
7. Protect Data and Privacy
We use data only for purposes permitted by the applicable consent, authorization, or agreement, and we protect it through privacy and governance measures appropriate to the information, its use, and our role.
Secondary use of data, including the development of AI models and the generation of real-world evidence, is bounded by what the originating consent, authorization, or agreement permits. We do not extend the use of data beyond those limits.
In practice: how a particular Oku company handles personal information is described in the privacy notice or Notice of Privacy Practices issued by that company.
Principles That Guide Our Work
These principles apply across Oku and inform the governance, policies, and practices through which we develop and use AI. We review our approach as the science, technology, and regulation of AI develop. We review these principles at least annually, and we update them as our business, the science, and the applicable regulation develop.
Responsibility for our AI governance framework sits with a named function within Oku, which maintains the policies, review processes, and documentation that give effect to these principles.
How These Principles Relate to Our Other Documents
These principles describe our commitments and our governance. They are not part of, and do not modify, any privacy notice, Notice of Privacy Practices, product documentation, or customer agreement. Information about how a particular Oku company handles personal information is provided in the privacy notice or Notice of Privacy Practices issued by that company. Product-specific information about intended use, performance, and limitations, and any disclosure required by applicable law for a specific product or market, is provided in the applicable product labeling and regulatory documentation.
About the Term AI
In these principles, an AI system means a machine-based system that, for explicit or implicit objectives, infers from the input it receives how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments. This is the definition published by the Organisation for Economic Co-operation and Development and reflected in international AI regulation.