In the past, the exchange of private information relied on secretly passing physical keys or sending messengers. Things changed with the introduction of cryptography. Today, many associate it with secure data communications, otherwise known as unilateral security. Even for the unaware, its functions of data confidentiality, authenticity and non-repudiation play a large role in everyday life.
However, there are still gaps to address. There is a need to consistently redefine the standards of cryptography as technology continues to evolve. Increasingly, there are more applications for a newer, more robust technology termed secure multi-party computation (MPC).
Professor Ronald Cramer, head and founder of the Cryptology Group at CWI Leiden University, discussed the matter in greater detail at a TECH TALK session on October 25, sharing with the audience its techniques and applications in today's world.
He defined MPC as a form of multilateral security, where two or more parties who want to process "mutually private data" in a way that allows them control over what kind of information should be released to the other.
"It's not a matter of who can be trusted or who cannot be trusted – there are many scenarios where there's mutual, conflicting interests, like in business," he said.
"And for this reason, people don't want to exchange everything they know."
To tackle the problem, the information thus has to go through a trusted and incorruptible arbiter, one that an MPC engine is capable of emulating virtually. It would compute functions from both sides and release only the results, leaving the rest of the data private. Professor Cramer likened the scenario to playing poker – a fair game between two players is only possible in the presence of a dealer to shuffle the cards.
"Suppose there's a world where there's only two entities – multilateral security still has meaning in a world of two because it does not need a third person to act as a mediator," he said.
There are already several real-life applications within industries, such as elections, auctions and Internet micro-auctions. Moving forward, Professor Cramer expects MPC to take on roles in benchmarking, machine learning, research data mining and distributed security.