Abstract: A method of maintaining a pool of network decoys includes evolving, over a plurality of epochs, the pool of network decoys towards one or more preferences of a network attacker. The method includes modeling preferences of the network attacker based on monitored interactions between the network attacker and the pool of network decoys to generate a preference model of the network attacker. Each epoch includes updating a fitness function based on the preference model and applying the fitness function to each network decoy included in the pool of network decoys to determine a plurality of fitness values, where the fitness values are representative of an attractiveness of a respective network decoy to the network attacker. The pool of network decoys is then updated based on the fitness values.

Abstract: A method and apparatus for mitigating row hammer attacks is provided. A row hammer alert is generated by a component of a memory architecture controlling operation of a memory device. The component may be a memory controller, coherency logic, or data fabric. The component obtains a physical address of an aggressor row that caused the alert and obtains an identifier of an execution context corresponding to the physical address. The component generates an error message for a processing device, the error message including the identifier of the execution context. The processing device retrieves the error message when performing a context switch. The processing device then generates an event received by the operating system. The operating system then takes action to reduce row hammer by the execution context, such as ending, restarting, or throttling the execution context.

Abstract: Mechanisms are provided for increasing the speed of an encoding process. The mechanisms classify operations in the encoding process based on learned associations between patterns of input data and corresponding encoding result types generated from encoding the patterns of input data. The mechanisms identify operations in the encoding process that can be pruned based on the classifications, to thereby generate a set of prune operations. In addition, the mechanisms replace operations in the set of prune operations with replacement operations that retrieve a corresponding previously generated result. Moreover, the mechanisms emit optimized encoding code comprising the replacement operations in replacement of the set of prune operations that are pruned, for execution of the encoding process using the optimized encoding code.

Abstract: An example method includes receiving an identifier associated with a security deficiency, wherein the security deficiency is associated with a computer system; determining, based on the identifier, text data associated with the identifier; determining a text prompt, wherein the text prompt comprises an instruction segment and the text data, and wherein the instruction segment identifies a remediation strategy detection task and an output constraint; providing the text prompt to a generative machine learning model; receiving, from the generative machine learning model, a set of outputs including a first output identifying a first remediation strategy and a second output identifying a second remediation strategy; determining that the first output satisfies the output constraint; determining that the second output fails to satisfy the output constraint; determining, based on the first output, a final output; and providing the final output using an output interface.

Abstract: Embodiments of this specification provide methods and apparatuses for data privacy protection. An embodiment of the methods comprises receiving, by a first party from a second party, an encrypted integrated vector, determining an encrypted result vector based on the original matrix and the encrypted integrated vector, determining a data processing result based on the encrypted result vector, and sending the data processing result to the second party for the second party to obtain a multiplication calculation result of the original matrix and the n original vectors based on the data processing result.

Abstract: A method and/or computer software for estimating the probability that a software weakness will be used in an exploit and/or malware and the probability that the developed exploit and/or malware will result in a compromise.

Abstract: A data file encryption transmission/reception system includes a computer with a registration module that encrypts input data files. The registration module includes a string conversion module that configured to convert the data file to be encrypted to string data, a key generation module that configured to generate a common key, and to divide the common key and the string data into first input information and second input information, a common key encryption module is configured to encrypt the first input information with the common key, to thereby generate common key encrypted data, a searchable encryption module that configured to encrypt the second input information in accordance with searchable encryption with a user secret key set in advance, to thereby generate searchable encrypted data and a communication module is configured to register a set of the common key encrypted data and the searchable encrypted data in the data management server.

Abstract: Techniques for managing the implementation of application-code scanning processes are disclosed. A system scans application code by analyzing metadata associated with the application code to identify a set of data needed to scan the application code with a scanning application. Based on the information obtained from the application metadata, the system identifies extraction processes that are needed to obtain the set of data. The system applies a set of one or more application-code scanners by implementing the extraction processes. The system presents in a graphical user interface (GUI) a set of results from scanning operations.

Abstract: A method includes obtaining a plurality of representative vectors associated with face-related data. The method includes determining an encryption key based on a parameter stored in a record, generating an encrypted vector set by, for each respective vector of the plurality of representative vectors, encrypting the respective vector with a homomorphic encryption operation based on the encryption key, where the encrypted vector set includes a first encrypted vector that is linked to a subset of the face-related data associated with the first plurality of face vectors. The method further includes obtaining an encrypted face search vector using the encryption key to perform homomorphic encryption. The method further includes selecting a first encrypted vector based on the encrypted face search vector and retrieving the subset of the face-related data based on the first encrypted vector.

Abstract: Cyber intrusion detection (CID) and data protection (DP) are coordinated within a storage node to enable the capabilities of DP to be automatically and quickly utilized in response to detected threats to help protect data. CID sends an alert message to DP in response to detection of a ransomware attack or other threat. DP responds to the alert message by implementing at least one countermeasure, such as: generating new targetless snapshots of the storage objects under attack, the version data group of which the storage objects are members, or all storage objects maintained by the storage array; securing and/or preserving some of the targetless snapshots that existed before the infection; changing the targetless snapshot generation and retention schedule; and temporarily halting generation of new targetless snapshots.

Abstract: Methods and systems for assessing risk associated with Artificial Intelligence (AI) models are provided. The method includes receiving a risk parameter among a plurality of risk parameters used for assessing an AI model, from a user associated with the AI model. Further, the method includes generating a preset questionnaire including a plurality of questions corresponding to the risk parameter. Furthermore, the method includes receiving user response data and corresponding evidence data against each of the plurality of questions from the user. The user response data is validated by correlating the user response data with the corresponding evidence data. The validated user response data is compared with preset response data corresponding to each of the plurality of questions. Thereafter, the method includes generating a report including a risk score indicating a risk associated with the AI model corresponding to the risk parameter, using a learning model, based on the comparison.

Abstract: A system and method for a continuous self-improving loop for security of generative artificial intelligence (GenAI) deployed in a system is provided. Prompt(s) are input to a target GenAI to perform security tests. An attack vulnerability report can be generated based on the output generated by the target GenAI in response to the plurality of security tests to a validation module that determines which data in the attack vulnerability report is accurate. For each security test that produced inaccurate output, a first dataset can be generated and can be used to create new security tests. For each breach of security that is accurate, a second dataset can be generated based on the respective security test that caused the breach of security and the target GenAI can be updated by training the target GenAI with the dataset to improve security of the target GenAI in use by the system.

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for facilitating modification of components of artificial intelligence computing applications via aggregated risk scores of individual artificial intelligence computing application components. The disclosed system generates data objects representing the artificial intelligence computing application components. The disclosed system determines mappings between the data objects based on relationships of the artificial intelligence computing application components. Furthermore, the disclosed system generates risk scores for the data objects representing the components of the artificial intelligence computing application by administering risk assessments that correspond to a system requirements framework.

Abstract: Providing a risk analysis report for an undesired event includes receiving a request comprising undesired event data defining an undesired event, wherein the undesired event data corresponds to an asset, identifying, from an attack database, one or more sets of attack attributes related to the undesired event data, and identifying one or more protection measures relevant to the asset and the attack attributes. The technique also includes performing a first set of Monte Carlo simulations for the undesired event based on combinations of the attack attributes, the protection measures, and the asset, and performing a second set of Monte Carlo simulations for the undesired event based on a subset of the combinations of the attack attribute. One or more combinations of the attack attributes, the protection measures, and the asset which result in the undesired event are provided in accordance with outcomes.

Abstract: The present disclosure provides techniques for processing a three-dimensional (3D) object file in a privacy-preserving manner. An example method includes obtaining an object file that comprises a specification of a 3D printable object, encrypting the object file using a public key to generate an encrypted object file, and sending, to a remote computing system, the encrypted object file and a request to process the encrypted object file to identify a characteristic of the 3D printable object. The method also includes receiving an encrypted result file from the remote computing system, wherein the encrypted result file comprises an encrypted Minkowski sum of the encrypted object file and an encrypted comparison file. The method also includes decrypting the encrypted result file using a private key corresponding with the public key to generate an unencrypted result file and processing the unencrypted result file to determine the characteristic of the 3D printable object.

Abstract: An object movement method includes receiving a first input of a user, where the first input is used to move a target object in a first space into a second space; and in response to the first input, moving the target object into the second space in a case that the second space is enabled and that a size of a residual space of the second space meets a predetermined condition, where the second space and the first space are independent of each other.

Abstract: A request to activate a service may be received from a user device and a determination may be made as to whether the request is authorized or fraudulent. In particular, a geographical location of the user device may be determined. Whether to activate the service may be determined based on the geographical location of the user device.

Abstract: The retention lock (RL) status for a backup file stored in a storage target is certified by obtaining the RL status and encrypting it using an encryption key process to create a certified RL status. This signs the RL status by the entity storing the backup file, rather than an application setting the retention lock. The certified RL status is provided as a token to backup software of a deduplication backup system, wherein it can be made available for inspection and audit. The data may include opaque data that is data not interpreted by the filesystem. The request for RL status includes the opaque data, which is returned as part of the response, and which can be returned in part as cleartext.

Abstract: Techniques for cybersecurity analysis. A method includes identifying a first set of paths to a first asset. The first set of paths includes a first path which allows for uploading inspection code, a second path which allows for running the inspection code, and a third path which allows for obtaining results of running the inspection code. The first set of paths is selected such that application programming interfaces (APIs) of the first set of paths are at least partially shared with APIs of a second set of paths to each of at least one second asset and, further, selected based on at least one request processing attribute of the at least one API used in order to access the first asset. The inspection code is uploaded via the first path and run via the second path. Outputs of the inspection code are obtained via the third path and analyzed.

Abstract: The retention lock status for a backup file stored in a storage target of a deduplication backup is certified by obtaining the retention lock status and encrypting, in the storage target, the retention lock status information using an encryption key process to create a certified retention lock status. This signs retention lock status by the entity storing the backup file, rather than an application setting the retention lock. The certified retention lock status is provided as a token to a backup software of the deduplication backup system, wherein the retention lock status information is made available for inspection and audit. It is then returned, in response to an audit request, for comparison against a cleartext representation of changes to retention lock status.