Security of the data and algorithms and results of artificial intelligence takes new methods. At OODAloop.com Matt Devost provides a framework for improving the security and reducing the risk of problems from your AI deployment in a post titled:
His approach: Build a strategy informed by four key components:
- Secure your AI infrastructure
- Secure the algorithms
- Security the training data
- ID and manage external risks and dependencies
His recommendations also underscore the importance of a strategy to drive results in each of these areas.
This premium OODAloop content was unlocked by Cooley LLP.
- OODAcast– A Conversation with Dan Gerstein - April 1, 2020
- Update on The End Coronavirus Project and Need for Volunteers - March 28, 2020
- Think Twice Before Deciding To Use A Personal VPN: You could be getting some really bad advice - March 23, 2020