Assurance Resource Providers
Description
Assurance Resources help accelerate the market adoption of AI models by establishing processes and tools developers can use to aid in streamlining and accelerating the adoption of AI in healthcare. CHAI is tasked with developing an evaluation framework for AI solutions in healthcare using consensus-driven standards and best practices.
Timeline
TBA
To help developers to trust these assurance resources, CHAI is certifying assurance resource providers who meet the following standardized criteria in the following areas:
1. Disclosing and minimizing conflicts of interest;
2. Protecting developer’s intellectual property;
3. Ensuring standardized disclosures and processes for data controllers/processors;
4. Demonstrating that the data provided to the model developer meets the data specification /requirement required by the model developer;
5. Ensuring that the model’s performance report meets the defined parameters approved by the model developer; and
6. Optional publishing of objective performance metrics using a CHAI model card on the CHAI National Registry.
As such, these resource providers will play a critical role in establishing trust in AI technologies by providing transparency of model performance within specific environments and by providing AI users and beneficiaries with more information about how this technology was evaluated and how it performed in other environments. These Assurance Resources are not part of any government regulatory process and are purely led by the private sector.