What is your role in the federal identity community?
Before joining NIST almost six years ago, I devoted more than 20 years to biometric systems engineering, technology assessment, and standards development. Since joining NIST, Artificial Intelligence (AI) and Machine Learning (ML) have been applied increasingly to advance biometric matching algorithms, and so I became interested in understanding more about how such approaches might impact biometric algorithm and system performance, and deployment. Currently, I am exploring how US and global approaches and guidelines for the ethical and trustworthy use of AI/ML technology can be applied to the measurement and understanding of biometric identification technology.
How does FedID help you succeed?
FedID offers an opportunity to network with identity community colleagues from across government, industry, academia, and the public to share the results of our work and maintain awareness of the latest technical advances and challenges, applications, and concerns. From an agency perspective, such interaction provides added insight into how NIST can best fulfill its mission to conduct independent and quantifiable science-based assessments to meet the needs of policy-makers, agency officials, and the broader identity community. Such engagements also help us to identify and better support community needs for new or updated standards and best practices.
What are you most looking forward to at FedID?
I am looking forward to the resumption of in-person interaction with colleagues from across the identity community and learning of their technological advancements, and new and enhanced ways they are applying identity technology to improve security, privacy, equity, and quality of life.
What advances in biometrics are you hoping to see at FedID this year?
It is always exciting to see improvements in biometric sensor technology, particularly with regard to such factors as speed, quality, size, and operating environment. Given the rapidly increasing use of AI to improve matcher and end-to-end system performance, I am hoping to see a significantly increased focus on and advances in the evolution and implementation of best practices for the trustworthy use of AI across product and application life cycles. This would include approaches to minimize, monitor for, and mitigate potential biases, both algorithmic and human; and maximize operational transparency and stakeholder confidence. The global community of technologists, standards developers, and policymakers have been working to this end across all applications of AI. The NIST AI Risk Management Framework currently under development should be a key resource in this regard as well.