Growing demand for a platform to enable secure collaboration between data stewards and algorithm developers spurs application of zero trust, confidential computing in healthcare
BeeKeeperAI, a pioneer in zero trust, confidential computing and privacy-preserving analytics for the training, validation, and deployment of artificial intelligence (AI) in healthcare, today announced its expanded collaboration with Microsoft (Nasdaq: MSFT) to allow algorithm developers and healthcare data stewards to access BeeKeeperAI, operating in the Microsoft Azure confidential computing (ACC) environment, through the Azure Marketplace.
The enhanced collaboration aims to accelerate the understanding and use of confidential computing as the new frontier for healthcare AI research, development, deployment, and monitoring. BeeKeeperAI utilizes the security and scalability of the ACC infrastructure within its applications and services to create a secure and confidential collaboration environment where proprietary AI algorithms can be validated against privacy-protected data sets from multiple organizations. A zero-trust confidential compute environment is designed to prevent visibility or access of the data or algorithm models by algorithm owners, data stewards, and BeeKeeperAI. This ensures data confidentiality and protects intellectual property.
“Through our collaboration with Microsoft and our use of Microsoft Azure services, we have developed key insights that will help to prepare the healthcare AI marketplace for confidential computing and provide resources to help data stewards and algorithm developers to effectively work in this new environment,” said Michael Blum, MD, Co-founder and CEO of BeeKeeperAI. “We chose Azure because of its privacy and security for data and algorithms as well as the infrastructure needed to compute at scale, in the cloud, across multiple locations.”
“The opportunity for AI to enable the delivery of better healthcare outcomes continues to expand exponentially, but developers are limited by access to critical datasets to train and to deploy their algorithms. We are pleased to partner with BeeKeeperAI to help the healthcare industry develop the understanding and expertise it needs to leverage confidential computing within healthcare innovation,” said John Doyle, Global Chief Technology Officer, Healthcare & Life Sciences at Microsoft. “The impact of what we are enabling aligns with the goals of the ever-increasing number of developers who are creating new ways to help to detect, diagnose, and treat diseases.”
To secure regulatory approval and clinical adoption, algorithm developers have to prove the efficacy of their models but have difficulty accessing patient datasets without exposing their intellectual property or compromising patient privacy. To address this challenge, BeeKeeperAI has implemented its platform for validating algorithms on protected data sets maintained in the Azure HIPAA compliant cloud environments of each participating data steward.
The Azure confidential computing environment includes Intel® SGX secure application enclaves with minimal cyberattack surface. The keys for the encrypted data and model to move into the enclave for computing, allowing data stewards to keep sensitive patient data private while simultaneously enabling algorithm developers to test their encrypted models against the encrypted data. Through Azure, algorithm developers can access information from multiple data stewards to rigorously validate the generalizability of their algorithms.
BeeKeeperAI is now offering its insights on how to successfully validate AI within the confidential computing environment through its new workshop offered in the Azure Marketplace. The workshop addresses both the algorithm and data preparation, encryption, and uploading into the cloud environment (where they remain), ensuring that they are protected throughout the compute process.
For more information about the BeeKeeperAI workshop, please visit the Azure Marketplace.
For more such updates and perspectives around Digital Innovation, IoT, Data Infrastructure, AI & Cybersecurity, go to AI-Techpark.com.
