Systems, methods, and computing device readable media for implementing fast oblivious transfer between two computing devices may improve data security and computational efficiency. The various aspects may use random oracles with or without key agreements to improve the security of oblivious transfer key exchanges. Some techniques may include public/private key strategies for oblivious transfer, while other techniques may use key agreements to achieve simultaneous and efficient cryptographic key exchange.

A method for performing spacetime-constrained oblivious transfer between a party A and a party B. The method includes imposing relativistic signaling constraints on a cryptographic task of one out-of-m oblivious transfer involving parties A and B. The method further includes using quantum systems for the one-out of-m oblivious transfer. The method guarantees unconditional security of the spacetime-constrained oblivious transfer, based on the imposed relativistic signaling constraints and based on using quantum systems for the one-out of-m oblivious transfer.

A method, apparatus and computer program product to detect whether specific sensitive data of a client is present in a cloud computing infrastructure is implemented without requiring that data be shared with the cloud provider, or that the cloud provider provide the client access to all data in the cloud. Instead of requiring the client to share its database of sensitive information, preferably the client executes a tool that uses a cryptographic protocol, namely, Private Set Intersection (PSI), to enable the client to detect whether their sensitive information is present on the cloud. Any such information identified by the tool is then used to label a document or utterance, send an alert, and/or redact or tokenize the sensitive data.

Implementations of this specification provide methods and apparatuses for oblivious data transfer between computing devices. An example method includes receiving, by a second computing device, an oblivious transfer from a first computing device. The first computing device splits feature data in a feature dataset into a plurality of sub-data and uses the plurality of sub-data as input, and the second computing device uses label data in a label dataset as input. The second computing device selects target sub-data from the plurality of sub-data input by the first computing device, and determines a first summation result of the selected target sub-data. The second computing device receives from the first computing device a second summation result of the one or more splitting parameters in the splitting parameter set, and calculates a statistical indicator based on the first summation result and the second summation result.

An item rating and recommendation platform identifies rating data including respective ratings of multiple items with respect to multiple users; identifies user-feature data including user features contributing to the respective ratings of the multiple items with respect to the multiple users; and receives, from a social network platform via a secret sharing scheme without a trusted initializer, manipulated social network data computed based on social network data and a first number of random variables. The social network data indicate social relationships between any two of the number of users. In the secret sharing scheme without the trust initializer, the social network platform shares with the item rating and recommendation platform manipulated social network data without disclosing the social network data. The item rating and recommendation platform updates the user-feature data based on the rating data and the manipulated social network data.

A computing device includes an interface configured to interface and communicate with a communication system, a memory that stores operational instructions, and processing circuitry operably coupled to the interface and to the memory that is configured to execute the operational instructions to perform various operations. The computing device processes an input value in accordance with a Threshold Partially-Oblivious Pseudorandom Function (TP-OPRF) blinding operation to generate a blinded input. The computing device then selects a threshold number of shareholder computing devices that are associated with a Key Management System (KMS) service and transmits the blinded input to them. The computing device then receives at least a threshold number of blinded output components from at least some of the shareholder computing devices and processes them to generate a blinded output. The computing device then processes the blinded output in accordance with a TP-OPRF unblinding operation to generate a key.

A computing device implements a key management system (KMS), and includes an interface, memory, and processing circuitry that executes operational instructions to maintain structured key parameters and a generating procedure associated with associated with a structured key. The generating procedure produces the structured key from an Oblivious Pseudorandom Function (OPRF) output, and the structured key parameters. The computing device receives a blinded value associated with the structured key from a requesting computing device, processes the blinded value using an OPRF secret to generate a blinded OPRF output, and returns the blinded OPRF output, the generating procedure, and the structured key parameters to the requesting computing device, which uses that information to generate the requested structured key.

A computing device including a processor, memory, and instructions, interfaces with a key management system (KMS) that provides encryption keys using an Oblivious Pseudorandom Function (OPRF). The device obtains, based on a type of encryption key being requested, a public key of a public-private key pair. The device creates an Oblivious Key Access Request (OKAR), including a blinded value associated with a requested encryption key. The OKAR is transmitted to the KMS, and a response is received. The response includes a blinded OPRF output, which yields an OPRF output as a result of being subjected to an unblinding operation. The OPRF output is validated using the public key, either directly or via a challenge, and in response to a positive validation, the OPRF output is used as a final key, or an intermediary key used to derive the final key.

Some embodiments are directed to a data retrieval device 210 for data-obliviously copying a subarray of a first array to a second array. The length of the second array is more than one and less than the length of the first array. The length of the subarray is at most the length of the second array. For each first element at a first index in the first array, the data retrieval device selects a second index in the second array for the first index in the first array; data-obliviously computes a choice bit indicative of whether to copy the first element to the second index in the second array; and replaces a second element at the second index in the second array by a replacement element, the replacement element being data-obliviously set to the first element or the second element based on the choice bit.

Embodiments are directed towards classifying data. A machine learning (ML) engine may select an ML model that may employ a cryptographic multi-party computation (MPC) protocol based on model preferences, including a parameter model, provided by a client. A randomness engine may be employed to provide random values and other random values based on the MPC protocol such that the random values may be provided to the client and the other random values may be provided to an answer engine. Input values that correspond to fields in the parameter model may be provided by the client such that the input values may be based on the MPC protocol and the random values. The answer engine may be employed to provide partial results to the question based on the ML model, the input values, and the MPC protocol that may be provided to the client.