Zero-Knowledge Proofs as a Service

[jonadecker]

Project description

Technologies based on zero-knowledge proofs have the potential to allow government digital services to offer privacy safeguards that are unmatched by the private sector, while simultaneously being easier for government teams to deploy, with respect to legal and compliance burdens.

However, these technologies are not fully commercialized and are currently difficult to use. This is an exception to the almost-always-true digital services rule that says "the technology isn’t the hard part".

Zero-knowledge proofs are complex, but they deliver what their name implies. For example, they can allow two people to mathematically prove to one another which one is older without revealing their ages, or any other information. In many cases, this can eliminate the need for government data-sharing (and the burden of risk that comes with it).

Many interactions between agencies that traditionally involve sharing data are actually questions of proof. Often, they do not actually require the transfer or storage of the underlying data at all. ZK-proofs have already permitted government agencies to deliver value to the public in ways that were previously declared impossible because of the risks and liabilities imposed by less-private technologies, such as PPRL. After this real-world discovery, the project team decided to pivot from the phase 1's "PPRL as a Service" to phase 2's "ZK-proofs as a Service".

This 10x Phase 2 project aims to assess the current feasibility of using zk-proof-based technologies in government digital services, and to deliver demonstrative software to aid in their adoption.

Was: PPRL as a Service

An idea submitted by a civil servant at: The General Services Administration, Federal Acquisition Service (FAS)

[jimmoffet]

Update: Meeting with NIST next week to discuss reviving Census xD's plan for a broadly-shared sandbox of hands-on PETs demos hosted on cloud.gov.