Shafi Goldwasser talks cryptography, cloud computing at Grace Hopper conference
The Anita Borg Institute’s Grace Hopper Celebration of Women in Computing conference started off with a bang Wednesday morning with a keynote speech by Massachusetts Institute of Technology Professor Shafi Goldwasser.
Goldwasser, who has received a wide array of awards for her work in computer science and cryptography, took the stage at the Phoenix Convention Center to discuss challenges in encryption, cloud computing and security.
The goal of her address was to present a “new cryptographic lens” to the audience, which could help show a new way for cryptographers to think about addressing modern problems, she said.
The main difference between modern and past cryptography is that modern cryptographers are not as focused on the wartime uses of their creations, Goldwasser said.
“Cryptography today is not just about fighting the bad guys,” she said. “It’s really about correctness and the privacy of computation.”
Goldwasser said the next generation of cryptographers will have to solve the problem of maintaining enough security that malicious users or programs cannot see certain information while still allowing any necessary data to be used.
The practical application of this idea could work its way into cloud computing systems, which people increasingly use for storing private information, she said. Cloud storage gives users a lot of flexibility, but it opens their data up to outside attacks or surveillance.
“The question is, can mathematics and technology help us to have the best of both worlds?” Goldwasser said. “We can, to a large extent, with the help of some sort of regulation, get most of the benefits without as much of the risk.”
Goldwasser used the examples of law enforcement conducting an investigation or hospitals sharing specific medical information to demonstrate how individual pieces of information can be encrypted separately. This maintains the privacy of the user and the security of the rest of the data while opening up specific pieces of information at will.
The other advantage of this approach is the ability to allow programs to analyze encrypted data and produce an encrypted answer or file, she said. Previously, encrypting information meant sacrificing a certain amount of usability.
“It turns out that many things that seemed impossible in the past are capable using some cryptographic thinking,” she said. “I believe that a lot of the challenges for the future of computer science are to think about new representations of data. And these new representations of data will enable us to solve the challenges of the future.”