DYNAMIC AND DELEGATED SHARED SECRET CRYPTOGRAM SCHEME

Description

The present invention involves the development of a novel cryptographic method (based on homomorphic encryption, proxy re-encryption, and shared secret techniques) that allows encrypted data to be stored in any type of public storage and subsequently decide how and where that data can be consumed, providing flexibility without compromising confidentiality. To this end, the owner of the encrypted data can authorize an intermediary to adapt the type of encryption based on the specific use and recipient, maintaining data confidentiality throughout the process with respect to the intermediary or any other entity other than the end consumer. Furthermore, some forms of consumption (using advanced cryptographic techniques in the field of Secure Multiparty Computing) allow recipients to operate on the confidential data without needing to see the clear content.

Advantages

The main advantages are the following:
- Flexibility in the process of using confidential data.
o Storing encrypted data at the time of generation and
making the decision on its use later.
o Different forms of consumption, such as owner recovery,
delegating decryption to an authorized party, or processing
with confidentiality guarantees through a Multi-Party
Secure Computing committee.
- Reducing the burden of adapting data to the type of consumption and transmission on the part of the owner.
o The adaptation process is performed by the intermediary (but without access to the clear content).
- The decision to use the data can only be authorized by the owner, using secret cryptographic keys.

Uses and Applications

Any application that generates data that needs to be confidentially protected (permanently or temporarily) and that can be recovered and used in the future using various techniques. Some examples of applications could include:

- Calculation of statistics on the aggregate of data generated by smart electricity meters without exposing individual data
- Collaborative fraud detection process by multiple banking entities based on the aggregate of individual transactions
- Secure aggregation of machine learning models with a federated architecture, avoiding the exposure of partial models.