Responsible AI in Health Care: Workshop with Roche
TLDR: This January, we hosted a successful two-day workshop to help Roche leaders build an implementation plan for Responsible AI best practices and showcase their leadership in the healthcare field.
We facilitated a two-day hybrid workshop with over 25 Roche colleagues across diverse functions to build their Responsible AI (RAI) practice. We held this workshop in the space of their partner, Mila, in Montréal, Canada, where many global participants were able to meet each other physically for the first time. Workshop attendees represented Roche offices from across the globe and across departments, including Pharma, Diagnostics, Products, Informatics, and Regulatory.
RAI Institute’s Ashley Casovan and Alyssa Lefaivre Škopac facilitating a
conversation on a Responsible AI in Health Care case study.
The workshop objective was to have participants learn from each other and advance the mission, vision, and objectives for Roche’s RAI work and Community of Practice. This included developing and defining core RAI principles for Roche, clarifying the context of RAI standards, aligning workstreams within Roche, and building a prioritized roadmap to deliver on RAI principles and objectives.
At Roche, patients and science are at the heart of everything they do. With their combined strength in diagnostics and pharmaceuticals, they’re driving personalized healthcare forward and adding their expertise in new areas, like artificial intelligence, real world data collection and analysis and collaborating with different sectors and industries.
One of the reasons for its long track record of success is the ability to share best practices across the organization. With RAI becoming imperative, the RAI Institute facilitated this workshop at the request of Roche leaders.
"At Roche we recognize the tremendous capability AI brings to healthcare research and the delivery of care. Roche was founded on the basis of quality and reliability in our research and the products we develop to help patients, which carries through today in our approach to embracing AI technologies. RAI Institute has done tremendous work to help organizations evaluate the emerging regulatory and standards landscape for AI, and develop the controls that can ensure new AI technologies are safely and appropriately developed and managed."
- Mike Bales, Director of AI & Digital Health
Roche participants and Mila staff in Mila’s Montréal space.
On Day 1, the team discussed upcoming AI-related legal requirements and RAI principles. A panel kicked the workshop off with a fascinating discussion on “Responsible AI in Practice: Exploring the Possibilities,” featuring moderator Ashley Casovan and panelists Nadine Bienefeld, Senior Researcher in Human Factors and Human-AI Teaming at ETH Zurich; Randy Goebel, External Advisor on AI at the Alberta Machine Intelligence Institute; and Sally Sfeir-Tait, CEO of RegulAtion.
The panel discussed the need for oversight in developing AI systems for health care and life sciences in a responsible manner, focusing on balancing cutting-edge technology with patient and organizational protection, building trust in AI systems, and developing safe practices through small-scale, multidisciplinary planning and testing.
Participants listening to the panel conversation on Day 1.
On Day 2, our discussions of RAI principles and the RAI regulatory and governance ecosystem deepened. The day started with a talk from Mark Sendak of Duke University, who discussed his experience in putting principles into practice and accelerating innovations at the Duke Institute for Health Innovation.
Leaders across Roche’s functions shared knowledge through presentations on their relevant work. The group discussed key considerations of RAI and further refined the mission, vision, and objectives for the Roche Community of Practice.
Alyssa Lefaivre Škopac and Amanda Lawson from the RAI Institute.
A huge thank you to Roche for the opportunity to connect! Thank you especially to Mike Bales, Director of AI & Digital Health; Aimee Park, AI & Digital Healthcare Partner, Academic Collaborations and Programs at Roche Integrated Informatics; and Roche Canada for the support in shaping the workshop from end to end. And thank you to each of the participants for your time and participation in this critical work.
Our work with Roche is building on the RAI Institute’s work with helping health care organizations, large and small, implement RAI. If you’re interested in our support with your organization’s Responsible AI practice, please reach out to Alyssa Lefaivre Škopac at email@example.com