ChatGPT's OSS Malware Review - Henrik Plate, Endor Labs
Techstrong TV Interviews
•
01-Jan-1900
Endor Labs conducted an experiment into how well ChatGPT performs at OSS malware review. Henrik asked the AI to classify 1,800 binaries for artifacts published on PyPI and npm as either malicious or benign. His conclusion is large language model-assisted malware reviews are not yet a viable alternative to manual reviews, but they can be one additional signal and input. They can help automatically review larger numbers of malware signals produced by noisy detectors, but one inherent problem is the reliance on identifiers and comments to “understand” code behavior. They are a valuable source of information for code developed by benign developers, but they can also be easily misused by adversaries to evade the detection of malicious behavior.
Up Next in Techstrong TV Interviews
-
Balancing FinOps and Resource Optimiz...
It is more important for organizations to monitor and control costs in the cloud. Many have implemented FinOps as a way to manage compute resources. However, organizations face a paradox: to save money via FinOps, they have to spend more on the technology and the people running it. To mitigate th...
-
Celebrus and Consumer Data Privacy - ...
Bill and Alan talk about Celebrus and the importance of clear opt-ins and outs for consumers. They also discussed how companies are now gathering consumer information in the wake of the fall of third-party data, and the significance of no-party data as a new innovation in the data world, particul...
-
Brian Singer-TSTV
