Have you ever wondered if there is a possibility to compute the average salary of three people, who do not want to tell, how much they earn? Yes, of course, there is. Maybe it is not so obvious a solution, because to calculate this average, you need to know a bit of cryptography, but it is possible. This method’s name is multi-party computation and it is very useful while solving problems, which need a lot of data, but privacy has to be maintained.
WHAT IS MULTI-PARTY COMPUTATION?
Multi-party computation is a subgroup of cryptography, which helps create methods for solving problems where information must be exchanged, computation must be performed, but without trust in one another. These assumptions are achieved, because the algorithm makes input data split into pieces and masked. It is only possible, when input data is delivered from more than one source.
HOW CAN BUSINESSES USE IT?
It sounds complicated, isn’t it? In this case, let’s take a look at a short example, how multi-party computation can be used in business. Nowadays, each bank analyzes data from transactions within their own IT. However, in the global world frauds are organized in a much more complicated way, so that it is hard to predict and prevent them. They are often spread between banks, and that is the reason why they are difficult to detect. Applying multi-party computation would allow analyzing of transactions in different banks fully protecting privacy. That would help to detect complicated frauds, which are spread between banks.
BENEFITS OF MULTI-PARTY COMPUTATION
You already saw an example, how can it be used in reality. It shows how many possibilities this solution opens in front of us. What are the biggest benefits of it?
- Only you see your data – you do not need third-parties anymore, whom you have to trust, because their role is to keep your data safe. Your data stays with you, is never transferred outside your internet firewalls.
- Quantum-safe – In this technology you do not have to worry about quantum computing anymore. Because it is encrypted in-use, by breaking up data and distributing it randomly across different players, it is safe.
- General Data Protection Regulation conditions are met – the data never moves from clients servers, so GDPR requirements about data privacy are met. The same is with other regulations (like HIPAA (protecting our personal health information), GLBA (protecting our financial data) or PCI DSS (covering credit card data)).
- Eliminates tradeoff between usability and privacy – because all data stays by its owner, there are no barriers in using data in analysis, you do not have to mask or drop any features.
As you can see, multi-party computation is an amazing possibility, to protect our data or data of our customers and at the same time using information resulting from them. This technology can be used in different industries, it can help solve complex problems like traffic plans in big cities or predict health outcomes, which will result in better patient treatment. All of these maintain privacy. Shortly, multi-party computation allows companies to work together, without trusting one another.