In June 2019, NTT Research Distinguished Scientist Brent Waters was named a Simons Investigator in Theoretical Computer Science. Each year, the Simons Foundation supports theoretical scientists in the fields of Physics, Mathematics, Astrophysics and Computer Science and their home academic institutions with significant research support for five years. Dr. Waters is a Professor in the Department of Computer Science at the University of Texas at Austin and recipient of several prestigious awards, including the Association for Computing Machinery (ACM) Computer and Communications Security (CCS) Test-of-Time and Grace Hopper Awards.
In its announcement of the award, the Simons Foundation noted Dr. Waters’ pioneering work in the fields of attribute-based encryption and functional encryption. To learn more about Dr. Waters, who joined NTT Research as one of its first Distinguished Scientists this summer, we sent him a few questions. An edited transcript of this Q&A follows:
Are you on sabbatical or are you dividing your time between NTT Research and UT Austin?
I am currently on leave from UT Austin and working for NTT Research. Both positions are centered on performing basic research in cryptography, so from that perspective dividing up my time is not so difficult. For UT Austin, I am still supervising my current Ph.D. students. And at NTT Research I am working on other aspects, such as recruiting new people to the lab and helping its visibility. Balancing these two responsibilities is probably the biggest challenge.
What part of your previous work do you consider the most significant, both personally and in terms of contributing to the field of cryptography?
It is really hard to pick out my favorite piece of research. Early on in my career I introduced (with Amit Sahai) the concept of Attribute-Based Encryption, which was about rethinking encryption as more than targeting just a specific user. Then starting around 2007, we started to push the concept even further toward a notion now called Functional Encryption. (People often credit a 2011 paper I co-authored for this, but actually the concepts and the talks date back earlier.) I remember it was initially a challenge to convince people that rethinking encryption was a central and a fundamental problem. However, it is satisfying to see that every year now these areas are a major focal point of cryptographic research.
How does witness encryption fit into your research agenda?
Witness encryption is a fascinating problem. The idea is that instead of encrypting to a cryptographic key, I can encrypt to a problem or puzzle. For instance, consider a graph consisting of nodes connected with edges. One central question is whether such a graph can be three-colored. That is, can I assign the colors of either red, blue or green to each node in the graph with the restriction that any two nodes with an edge between them cannot be labeled with the same color? In witness encryption, one could encrypt a message to such a graph and produce a ciphertext. Someone can decrypt and receive the message if they know a three-coloring for that particular graph.
Right now, the concept of witness encryption is out there; however, we don’t have any fully satisfactory solutions. Ideally, we would like to prove security under a “standard assumption” such as the difficulty of factoring. This seems like a rather challenging problem. You might characterize it as one of the more ambitious problems in the agenda.
Why did you decide to join the NTT Research CIS Lab?
The CIS lab seemed like a neat opportunity to do something new with my research life and bring several strong people together. One important factor is that we are dedicated to keeping the focus on basic or pure research.
As a Distinguished Scientist, what are you tasked with and hoping to accomplish?
As part of the Distinguished Scientist position, I am hoping both to be able to build the lab (i.e. hire strong people) as well as raise the visibility of the lab. However, I still want to focus mostly on being able to perform strong research myself. One challenge will be to keep this up while working into the role.