Google LLC’s cloud team today announced the release of a new solution for secure infrastructure and data processing called Isolator that adds protection for healthcare data and allows trusted collaboration between parties.

Taylor Lehmann, director of the Office of the CISO and enterprise health security officer at Google Cloud, told SiliconANGLE in an interview that one of the biggest problems that users in the healthcare industry face is dealing with regulated information, such as protected healthcare information, or PHI. The same is true of any other highly regulated industry such as banking, as well as for Social Security numbers, financial information or any sensitive data.

The problem becomes even bigger when that data needs to be shared with third parties for storage, analysis or building the next machine learning algorithm that might save lives. Patient healthcare data is sacrosanct when it comes to privacy, so organizations need to avoid exposing it to potential leaks or someone doing something malicious with it.

“There’s a whole spectrum of risks that go with sharing information with third parties,” Lehmann said and that includes Google. “We have tons of people who work here with extremely long experience in innovation and building transformer models, and the foundations of modern AI. We have a lot of companies that want to share data with us so that we can help them build new and interesting AI algorithms.”

To understand how Isolator works would be similar to the way that researchers work with a laboratory “isolator box,” or a glove box, that basically compartments an object away from the rest of the world so that nothing goes in and nothing comes out, and what’s inside can only be touched with rubber gloves. Similarly, Isolator is built in the cloud and third-party users can be granted access to data in it, but they cannot remove it.

Under the hood, Isolator is an environment built with BeyondCorp Enterprise, Google’s zero-trust technology that is coupled with other features available from Google Cloud, including VPC Service Controls and Chrome Enterprise. It also uses encryption at rest, in use and in transit. The system can also provide audit trails for anyone using the data with granular identity-based access so that if anything happens to the data at any time, it can be traced back to the person who manipulated it.

Isolator also works with even more powerful confidential computing capabilities such as Confidential Space, which supports encryption for data with direct hardware isolation and remote attestation for even greater control.

“Organizations don’t necessarily have to move their data around anymore,” Lehman explained. “They can keep it where it is, put type safeguards on it to make sure that essentially nothing truly bad can happen to it, while at the same time getting this work done in a highly collaborative way. In Google’s case, it’s helping us accelerate companies building the machine learning models, and taking advantage of this new technology without necessarily having to take on all that security risk.”

Lehmann said using Isolator opens up opportunities for researchers who are currently hampered by privacy and regulatory worries because they have data too sensitive to move so they don’t want to part with it. 

The new Isolator solution is available through Google Cloud’s Consulting Services. Customers can use Isolator to set up isolated instances within their information technology environments to work on sensitive healthcare data.

The solution can also be used to develop custom models using Google Cloud’s machine learning platform Vertex AI safely to train, deploy and customize AI and machine learning models and AI applications in isolation. It could be used with Google’s Multiomics Suite to simplify drug discovery for clinical trials or to scale analytic data processing using BigQuery.

Image: Everything Possible/Shutterstock