How Fast Do You Want to Go?

In the realm of PET (Privacy Enhancing Technologies), there exists a spectrum of tradeoffs that users must consider. These technologies offer distinct advantages and compromises, ranging from the degree of trust required from involved parties to computational speed and the diverse application spaces they cater to. Balancing these factors becomes pivotal when choosing the most fitting approach.

The near-term evolution of computing technologies like Cornami’s FracTLcore™ Compute Fabric, will significantly impact the usability and effectiveness of PET. Innovations such as Zero Knowledge Proofs, Privacy-Preserving Machine Learning (PPML), and Fully Homomorphic Encryption (FHE) represent strides toward bolstering privacy in different ways. Zero Knowledge Proofs allow validation of information without revealing the content, while PPML enables machine learning models to operate on encrypted data. FHE, in particular, promises ultimate security by adopting a “Trust no one” philosophy, though it comes at the cost of being the slowest and the highest in complexity among these techniques.

FHE stands out as a formidable approach due to its unparalleled security measures, emphasizing a robustness that trusts no intermediary. However, it’s important to acknowledge that this heightened security is coupled with slower operational speeds and increased complexity, factors that might not be ideal for every application or scenario.

Recognizing the strengths of individual PET techniques, the future is likely to witness combinations and synergies between them. For instance, the merging of techniques like Multi-Party Computation (MPC) with FHE or the integration of PPML with FHE could offer enhanced solutions that harness the strengths of each technique while mitigating their respective weaknesses. These combinations pave the way for more comprehensive and versatile privacy-centric approaches, catering to a wide array of needs and scenarios.