Microsoft's new health AI initiative may do some good for partner nonprofit and academic healthcare groups, but it may raise some data privacy concerns as well.
Microsoft launched AI for Health, a new $40 million philanthropic initiative that will see the tech giant provide certain health-related nonprofits and academic programs with AI and cloud computing tools, cash grants and collaboration with Microsoft's leading data scientists.
The initiative, revealed Jan. 29 while the world was still grappling with the global outbreak of the coronavirus, complements the broader work done by Microsoft in the healthcare industry, and is the fifth program under the tech company's AI for Good program, which provides similar support to a range of nonprofits and academic institutes.
The other programs under AI for Good are AI for Earth, AI for Accessibility, AI for Humanitarian Action and AI for Cultural Heritage.
AI in the healthcare industry
The use of AI in healthcare has grown rapidly over the last few years, with healthcare research groups and pharmaceutical companies using the technology to develop new drugs and train deep learning systems to improve imaging at scale and to quickly scan vast numbers of medical documents for related research or specific diseases or drugs, said Kjell Carlsson, senior analyst at Forrester.
"All of that has been opened up by AI," he said.
Microsoft's AI for Health initiative sees the tech company partner with nonprofit and research healthcare organizations.
Microsoft has partnered with organizations including the Novartis Foundation, Fred Hutchinson Cancer Research Center and its Cascadia Data Discovery Initiative, PATH, BRAC, St. Jude Children's Research Hospital, Intelligent Retinal Imaging Systems and Seattle Children's Research Institute, according to the Microsoft blog post about AI for Health.
The partnership with Seattle Children's Research Institute, for example, aims to further researchers' understanding of Sudden Infant Death Syndrome, still a largely unexplained phenomenon.
By combing through huge amounts of data, including data on millions of births and deaths, with data science tools in Microsoft Azure, Microsoft researchers found correlations that showed statistical increases in SIDS. They brought the information to Seattle Children's Research Institute and are working with the organization to discover more correlations.
AI in healthcare has many benefits, according to Carlsson, but he said he is unsure how the public will react to Microsoft increasing its AI healthcare efforts.
Privacy concerns
Recently, Google came under criticism after it was revealed that the tech giant had partnered with Ascension, one of the largest nonprofit health systems in the U.S. The partnership gave Ascension Google's cloud and AI tools, but also provided Google access to patient data. While appearing legal, the partnership was not immediately disclosed to the millions of patients whose records were accessed by Google.
With the backlash that was generated in November, Carlsson said it is somewhat odd that Microsoft, only two months afterward, decided to launch the AI for Health initiative.
The public tends to respond well to the benefits of AI in healthcare, such as more effective cancer detection or potentially lower healthcare costs, but tends to also express concerns about data privacy, according to Kjell Carlsson, senior analyst at Forrester.
The public tends to respond well to the benefits of AI in healthcare, such as more effective cancer detection or potentially lower healthcare costs, but tends to also express concerns about data privacy, he said.
Meanwhile, Microsoft noted the importance of data privacy in the AI for Health blog post, authored by John Kahan, the vendor's chief data analytics officer.
The company is working "to deliver advanced privacy technologies based on the work Microsoft and academic researchers invented years ago, called differential privacy, which makes it possible to extract useful insights from datasets while guaranteeing the privacy of individuals," according to the post. The differential privacy effort stems from collaboration between Microsoft and Harvard's Institute for Quantitative Social Science.
For healthcare organizations, a tech giant such as Microsoft can provide a single provider for cloud and analytics services and AI needs, Carlsson said. The vendor has a wide network of partner vendors on Azure, as well as its own tools, and has maintained a fairly good reputation for respecting data privacy, he said.
This is not the first time a major AI technology company has put funding into aiding the healthcare sector. Other tech giants, including Google, Facebook, and IBM, have poured funding and technology into academic, research and nonprofits for years.
