Sap Se

Endor Labs Inc

Sap Ag

Henrik Plate

Serena Ponta

Antonino Sabetta

Dimitrios Styliadis

Michele Bezzi

Francesco Di Cerbo

Alexandre Wilhelm

Cedric Hebert

Laurent Y Gomez

Andreas Schaad

Slim Trabelsi

Philip Hamer

Gilles Montagnon

Frederic Montagut

Mohammad Ashiqur Rahaman

Cedric S P Ulmer

Cedric R J Hebert

Joseph Hejderup

Elton Mathias

Theodoor Scholte

Georgios Gousios

Yves Roudier

Matteo Maria Casalino

In some examples, a server injects malicious code into a legitimate software package to create an injected package. The server uses an artificial intelligence to extract a plurality of parts from the injected package and to mutate individual parts of the plurality of parts to create mutated parts. The server assembles the mutated parts to create a mutated malware. A malware scanner determines a risk score associated with the mutated malware. Based at least in part on determining that the score satisfies a predetermined threshold, the server stores the mutated malware in a set of mutated malware and creates at least one additional mutation based on the mutated malware. After determining that a size of the set of mutated malware satisfies a requested size, the malware scanner is modified to increase detection of the malicious code in individual mutated malware in the set of mutated malware.

Using artificial intelligence to generate test data to test malware scanner(s)

In some examples, a server determines that project code calls a component in a third-party library that has been updated. The server determines a first type hierarchy associated with the component, a second type hierarchy associated with the updated component, and performs a type hierarchy comparison of the first and the second type hierarchy. The server determines a first call graph associated with the component, determines a second call graph associated with the updated component, and performs a call graph comparison of the first and second call graph. Based at least in part on the type hierarchy comparison and the call graph comparison, the server determines a set of breaking changes associated with the updated component, with each breaking change in the set identifying changes to the updated component, relative to the component, that are determined to cause an incompatibility when the project code calls the updated component.

Impact analysis of changes in third-party software components on downstream users

A non-transitory computer-readable media, method, and server for detecting and addressing vulnerabilities are described. For example, a server receives a security advisory of a vulnerability of a function in a third-party library, accesses a version control system of the third-party library to identify fix commits that address the vulnerability, determines files corresponding to functions in the third-party library that include the vulnerability, performs a comparison before and after a code change was made to the files, identifies functions that included the vulnerability and have been modified to address the vulnerability, performs a search for component versions that include the individual functions that included the vulnerability, generates an enriched vulnerability description that includes identifiers of package versions that include: fixed versions and vulnerable versions of the functions, and modifies project code in a development system to use the fixed versions of the functions.

Identifying and addressing potential vulnerabilities in third-party code

A non-transitory computer-readable media, method and server for detecting and addressing vulnerabilities in a third-party code are described. In some examples, a server receives a security advisory that includes a description of a vulnerability and accesses a version control system (VCS) used by a third-party library to determine additional resources related to the vulnerability. The server determines a set of code changes performed by the project maintainers in the VCS, identifies one or more fix commits that address the vulnerability, and identifies one or more functions with the vulnerability that have been changed by the fix commits. The server performs a search for components and component versions that include the one or more functions with the vulnerability and generates an enriched vulnerability description that includes identifiers of package versions that include fixed versions of the one or more functions and vulnerable version of the one or more functions. Project code in a development system is modified to use the fixed versions of the one or more functions.

Embodiments detect malicious code in distributed software components. A detector element references a source code repository (e.g., open source, commercial) containing lines of various files of a distributed artifact. Subject to certain possible optimizations, the detector inspects the individual files and lines of the artifact file-by-file and line-by-line, to identify whether any commit history information is available from a Versioning Control System (VCS). A risk assessor element receives from the detector element, results identifying those lines and/or files for which no VCS commit history is available. The risk assessor then references code features (e.g., file extension, security-critical API calls) in the results, to generate a probability of the malicious nature of the source code lacking VCS commit history information. An analysis report including this probability and additional relevant information, is offered to a user to conduct further manual review (e.g., to detect false positives attributable to benign/legitimate source code modification).

Detecting malicious components using commit histories

Systems and methods are provided for retrieving a set of code changes to source code from a source code repository, analyzing the set of code changes to generate a vector representation of each code change of the set of code changes, analyzing the vector representation of each code change of the set of code changes using a trained security-relevant code detection machine learning model, receiving a prediction from the security-relevant code detection machine learning model representing a probability that each code change of the set of code changes contains security-relevant changes, analyzing the prediction to determine whether the prediction is below or above a predetermined threshold, and generating results based on determining whether the prediction is below or above a predetermined threshold.

Security-relevant code detection system

Systems and methods, as well as computing architecture for implementing the same, for decoy injection into an application. The systems and methods include splitting a standard test phase operation into two complementary phases, and add new unit tests to the process, dedicated to testing the proper coverage of the decoys and avoiding non-regression of the original code.

Non-regressive injection of deception decoys

Systems and methods are provided for accessing a source code repository comprising a plurality of versions of code, analyzing the plurality of versions of code of the component to compute metrics to identify each version of code, analyzing the metrics to determine a subset of the metrics to use to as a fingerprint definition to identify each version of the code, generating a fingerprint for each version of code using the fingerprint definition, generating a fingerprint matrix with the fingerprint for each version of code for the software component and storing the fingerprint definition and the fingerprint matrix

Software version fingerprint generation and identification

Embodiments provide systems and methods configured to mine information available from informal sources (e.g., social media, blogs, and forums) regarding security vulnerabilities. Particular embodiments may comprise engine(s) of a backend in communication with a user through an interface of a frontend, and also in communication with an underlying database to store security information and related information (e.g. search parameters). Embodiments may allow creation of user-specific search phrases for searching information in one or more informal social media information sources. Search results may be consolidated, and users such as system administrators quickly alerted to possible security issues. Embodiments may refine data mining over time by tracking the reputation (e.g. for data accuracy, freshness) of various sources. Embodiments may also reference formal official and third party sources of security information.

Mining security vulnerabilities available from social media

Systems and methods are provided for accessing a source code repository comprising a plurality of versions of code, analyzing the plurality of versions of code of the component to compute metrics to identify each version of code, analyzing the metrics to determine a subset of the metrics to use to as a fingerprint definition to identify each version of the code, generating a fingerprint for each version of code using the fingerprint definition, generating a fingerprint matrix with the fingerprint for each version of code for the software component and storing the fingerprint definition and the fingerprint matrix.

Various embodiments of systems, computer program products, and methods for prioritizing software patches are described herein. In an aspect, the software patches are retrieved by querying software repositories. Further, code changes associated with the software patches are determined. One or more instances of bug fix patterns are identified in determined code changes. The software patches are classified based on the identified bug fix patterns. Priorities of the software patches corresponding to the identified instances of the bug fix patterns are determined based on the classification and a pre-defined policy. Upon determining priorities, the software patches are installed based on the priorities.

Prioritization of software patches

Automated systems and methods for assessing the urgency of installing a patch for a component of a software application are described. The systems and methods involve identifying a set of defective programming constructs of the component that are altered by the patch, collecting execution traces of programming constructs of the software application and programming constructs of the component in a context of application use, and evaluating the execution traces to determine whether one or more defective programming constructs of the component are invoked in the context of application use.

Software patch evaluator

Implementations are directed to enhancing assessment of one or more known vulnerabilities inside one or more third-party libraries used within an application program that interacts with the one or more third-party libraries. In some examples, actions include receiving a complete call graph that is provided by static source code analysis (SSCA) of the application program and any third-party libraries used by the application, receiving one or more stack traces that are provided based on dynamic source code analysis (DSCA) during execution of the application program, processing the complete call graph, the one or more stack traces, and vulnerable function data to provide one or more combined call graphs, the vulnerable function data identifying one or more vulnerable functions included in the one or more third-party libraries, each combined call graph being specific to a respective vulnerable function, and providing a graphical representation of each combined call graph.

Assessing vulnerability impact using call graphs

In accordance with aspects of the disclosure, systems and methods are provided for managing software component dependencies for applications using declarative constraint definitions, including enabling specification of constraint definitions using a declarative language to analyze and detect software component dependencies on one or more libraries that meet certain criteria, identifying and checking software component dependencies on the one or more libraries that violate the constraint definitions, and implementing an algorithm for analyzing applications and resolving software component dependencies on the one or more libraries that violate the constraint definitions.

Software dependency management through declarative constraints

The embodiments provide a data processing apparatus including a runtime check identifier configured to determine, in response to a request to execute a function of the application, whether a runtime check descriptor corresponding to the function is stored in a database and select the runtime check descriptor if the runtime check descriptor is determined as stored in the database. The selected runtime check descriptor provides, in part, at least one configuration check of a component related to an execution of the function. The data processing apparatus also includes a system component configured to invoke a system management tool to perform the at least one configuration check, and a runtime check controller configured to receive results of the at least one configuration check and perform at least one action specified in the selected runtime check descriptor based on the received results.

Growing community of inventors

Valbonne, France

Henrik Plate