### How to Generate and Use Universal Samplers

**Abstract:**
The random oracle is an idealization that allows us to model a hash
function as an oracle that will output a uniformly random string given any
input.
Weintroduce the notion of a universal sampler scheme that extends the notion
of a random oracle, to a method of sampling securely from arbitrary
distributions.

We describe several applications that provide a natural motivation for
this notion; these include generating the trusted parameters for many
schemes from just a single trusted setup. We further demonstrate the
versatility of universal samplers by showing how they give rise to simple
constructions of identity-based encryption and multiparty key exchange.
In particular, we construct adaptively secure non-interactive multiparty
key exchange in the random oracle model based on indistinguishability
obfuscation; obtaining the first known construction of adaptively secure
NIKE without complexity leveraging.

We give a solution that shows how to transform any random oracle into a
universal sampler scheme, based on indistinguishability obfuscation. At
the heart of our construction and proof is a new technique we call
"delayed backdoor programming" that we believe will have other
applications.

Joint work with Dennis Hofheinz, Tibor Jager, Amit Sahai, Brent
Waters and Mark Zhandry.