AI Security & Privacy at ClarityCare AI

Data Integrity and Ethical AI Implementations

ClarityCare AI is a tool for UM clinicians to use, to facilitate the intake of requests and determination of medical decisions. Prior to full implementation, we assess ClarityCare AI algorithm suggestions through retrospective reviews against nurse decisions to validate accuracy. After implementation, on an ongoing basis, ClarityCare AI conducts regular verification checks in order to maintain accuracy and address any emerging risks.

For cases that go through our algorithm, we run statistical tests on key variables every quarter (e.g. service lines). This ensures that there is no statistically significant difference in case outcomes for comparable cases across service categories.

ClarityCare AI does not make clinical decisions. It’s a tool used by clinicians to streamline their review process.

Ensuring Data Privacy and Integrity with AI

We never use our customer’s data for AI model training. Customer data collected is used solely for the purpose of performing the services in accordance with a customer’s services agreement. ClarityCare has robust security protocols in place to ensure data privacy, integrity and source verification. ClarityCare has multiple controls to prevent unauthorized data manipulation, including MFA and logs database.

‍

Vision on AI

Leveraging AI to enhance, streamline & expedite tedious data processing is the core of what we do. ClarityCare AI does not make clinical decisions. It matches criteria to clinical documents and makes recommendations. This accelerates the review process such that patients & providers get faster access to prior authorization outcomes. Our system is designed such that the default option is to defer to human expertise when there is any ambiguity. We don’t plan on fully automating the process as complex decisions must still be made by an expert.

‍

Our Trusted AI Partners

ClarityCare AI does not rely on a single AI system to function nor does it develop its own AI model. We are leveraging third party AI models and tailoring them to our use case. We have carefully selected our AI partners - OpenAI, Anthropic and AWS - for their expertise, innovative technologies and security protocols. These collaborations enrich our AI features, enabling us to provide our services and meet our customers’ varied and dynamic requirements.

Although we don't control the specific AI systems or third party API uptime, we have a mechanism in place to maintain functionality through alternative methods of accessing AI systems in the event that one of our partners is experiencing delays or downtime.

‍

Infrastructure & Data Transmission

All ClarityCare AI services are hosted on Amazon Web Services (AWS) in the us-east-1 region located in northern Virginia in the United States.

The setup of the data transmission process is straightforward. ClarityCare AI has a secure and reliable API through which data is transmitted to the customers’ endpoints. The process is highly secure & almost instantaneous.

‍

Regulatory Compliance for the AI Components

Certain state laws or regulations may limit AI from making certain medical benefit decisions (for example  SB1120 (CA), SB24205 (CO), HB2472 (IL) & A9149 (NY - not yet enacted)) Our system does not replace expert decision-making.  It makes recommendations based on all the information sent by the submitting provider and made available to ClarityCare AI. The recommendation process is entirely transparent and auditable.

ClarityCare AI services are compliant with HIPAA and other applicable federal regulations.   In addition, ClarityCare AI can empower organizations to become compliant with new federal mandates such as CMS 0057 F which requires new prior authorization API endpoints & documentation,faster turnaround time, transparent decision making trail & metrics.

‍

Conclusion

Please visit our Trust Center to learn more about our security and privacy policies.

You can reach out to us at security@claritycare.ai for more information. We will be happy to answer your questions.