Input / Output Privacy

What is the difference? 👀

Input Privacy:

Input privacy is a way of protecting data when different parties are sharing it for calculations or analysis. It ensures that each party can submit their data without revealing it to others. The goal is to keep everyone’s data private while still being able to work together on a common task.

Cryptographic protocols like Secure Multi-Party Computation (sMPC), Trusted Execution Environments (TEEs) and Homomorphic Encryption (HE) use encryption to perform calculations on sensitive input data.

💡 Tip: While offering theoretical guarantees, these PETs can be computationally expensive, especially with large datasets.

Output Privacy:

Output privacy, on the other hand, ensures that data shared outside the organization is protected from re-identification and misuse.

Output privacy lets you control what information you share with others. Just like you want to keep some details private when talking to friends, output privacy ensures certain information remains hidden in data analysis or sharing.

Techniques such as data masking, anonymization, and differential privacy help remove or obfuscate sensitive information while maintaining data utility for analysis.