 |
Just Released! Draft NIST SP 800-226, Guidelines for Evaluating Differential Privacy Guarantees & UK-US Privacy-Preserving Federated Learning Blog Series
Dear Colleagues,
We’re excited to announce the release of the NIST Special Publication (SP) 800-226 Initial Public Draft (IPD), Guidelines for Evaluating Differential Privacy Guarantees, which is all about differential privacy, a privacy-enhancing technology that quantifies privacy risk to individuals when their information appears in a dataset. In response to President Biden’s Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence, SP 800-226 is intended to help agencies and practitioners of all backgrounds—policy makers, business owners, product managers, IT technicians, software engineers, data scientists, researchers, and academics—better understand how to evaluate promises made (and not made) when deploying differential privacy, including for privacy-preserving machine learning. Additionally, there is a supplemental, interactive software archive that illustrates how to achieve differential privacy and other concepts described in the publication.
The comment period for this draft is open until 11:59 p.m. EST on Thursday, January 25, 2024. Visit our publication page for additional details about SP 800-226 and the comment form.
In addition, last week we launched a new blog series on privacy-preserving federated learning (PPFL) as a follow on to the past UK-US PETs Prize Challenges collaboration. Modeled after our successful differential privacy blog series, this joint UK-US series focuses on addressing the privacy challenges in federated learning, an approach that enables machine learning models to be trained across separate datasets. Over the coming months, we’ll be publishing a number of blogs to provide practical guidance on PPFL. The series will feature guest authors from organizations involved in the UK-US PETs Prize Challenges, and other leading experts in the field. Future topics will include privacy threat models in federated learning, solutions developed during the prize challenges, and resources for getting started with federated learning.
The UK-US Blog Series on Privacy-Preserving Federated Learning: Introduction | by Joseph Near, David Darais, Dave Buckley, and Naomi Lefkovitz: Read the post.
If you have any questions about:
- the SP 800-226 publication, please reach out by contacting privacyeng@nist.gov; or
- the blog series, please reach out by contacting pets@cdei.gov.uk and privacyeng@nist.gov.
Help us advance the adoption of PETs by providing feedback on these new releases!
All the best,
NIST Privacy Engineering Program