WPF urges National Science Foundation to study Statistical Parity

WPF’s Public comments on National Privacy Research Strategy

The World Privacy Forum submitted comments today to the National Science Foundation in response to its request for information about a national privacy research strategy. WPF urged a research focus on statistical parity and its implementation. Statistical parity is a term WPF’s Pam Dixon coined at the FTC’s Big Data, Tool For Inclusion or Exclusion? workshop in September 2014. Here is Dixon’s definition of the term:

“Statistical parity means ensuring that all parts of the consumer data analytics process are fair: data collection, which data factors chosen and used for analytics, accuracy of the factors and how well the algorithm works for its intended purpose, and then how the final results are vetted and used, and for how long. Statistical parity means finding ways to ensure privacy and fairness in the analytics process from beginning to end, and to ensure that decisions about consumers are accurate and used fairly and in a non-discriminatory way.” Comments to NSF, Oct. 17, 2014

In their comments, the WPF also asked for research on granular consent mechanisms and on consumer segmentation, scoring, and modeling.

WPF encouraged the study of data provenance and also study of layered protection models, a concept WPF has been researching. From the comments:

On layered protection models:

“We have hypothesized that a layered, overlapping protections approach is going to work best going forward. FIPs plus statistical parity, for example, working together as a model. Perhaps a responsible use framework could be added on top of this, in addition to, but not replacing the other frameworks. The days and era of having a single-silver bullet solution to privacy are long gone, and it is much more helpful to find effective ways of layering protections and building in individual consumer choice – meaningful choice – wherever possible. We would like to see a request for scenarios on this point. We emphasize, however, that any model that leaves out FIPs will be incomplete in approach.”

On meta tagging and data provenance:

“In some circumstances, it will be appropriate to tag data so that the source of the data is transferred whenever the data is transferred. In other words, we suggest that metadata be employed to show the provenance of the data. This is technically possible and already happens to some extent in Electronic Health Records, and. It has great potential for privacy. We observe in passing that better provenance of data has great value to researchers as well.”

To see more, read the full comments.

WPF’s Public comments on National Privacy Research Strategy (PDF)

Posted October 17, 2014 in Future of Privacy, Modern privacy, National Science Foundation, Public Comments, Statistical Parity

New Report: Risky Analysis: Assessing and Improving AI Governance Tools

We are pleased to announce the publication of a new WPF report, “Risky Analysis: Assessing and Improving AI Governance Tools.” This report sets out a definition of AI governance tools, documents why and how these tools are critically important for trustworthy AI, and where these tools are around the world. The report also documents problems in some AI governance tools themselves, and suggests pathways to improve AI governance tools and create an evaluative environment to measure their effectiveness. AI systems should not be deployed without simultaneously evaluating the potential adverse impacts of such systems and mitigating their risks, and most of the world agrees about the need to take precautions against the threats posed. The specific tools and techniques that exist to evaluate and measure AI systems for their inclusiveness, fairness, explainability, privacy, safety and other trustworthiness issues — called in the report collectively AI governance tools – can improve such issues. While some AI governance tools provide reassurance to the public and to regulators, the tools too often lack meaningful oversight and quality assessments. Incomplete or ineffective AI governance tools can create a false sense of confidence, cause unintended problems, and generally undermine the promise of AI systems. The report contains rich background details, use cases, potential solutions to the problems discussed in the report, and a global index of AI Governance Tools.
National IDs Around the World — Interactive map

About this Data Visualization: This interactive map displays the presence...
Report: From the Filing Cabinet to the Cloud: Updating the Privacy Act of 1974

This comprehensive report and proposed bill text is focused on the Privacy Act of 1974, an important and early Federal privacy law that applies to the government sector and some contractors. The Privacy Act was written for the 1970s information era -- an era that was characterized by the use of mainframe computers and filing cabinets. Today's digital information era looks much different than the '70s: smart phones are smarter than the old mainframes, and documents are now routinely digitized and stored and perhaps even analyzed in the cloud, among many other changes. The report focuses on why the Privacy Act needs an update that will bring it into this century, and how that could look and work. This work was written by Robert Gellman, and informed by a two-year multi-stakeholder process.