Feb 08, 2025Ravie LakshmananArtificial Intelligence / Supply Chain Security Cybersecurity researchers have uncovered two malicious machine learning (ML) models on Hugging Face that leveraged an unusual technique of "broken" pickle files to evade detection. "The pickle files extracted from the mentioned PyTorch archives revealed the malicious Python content at the beginning of … [Read more...] about Malicious ML Models on Hugging Face Leverage Broken Pickle Format to Evade Detection
detection
Using AI to Simplify Cloud Configuration Drift Detection
Cloud environments are dynamic by nature; they frequently change and update configurations. Unless these changes are carefully tracked, they could lead to configuration drift, a situation in which the runtime state of a resource deviates from its intended baseline configuration. Configuration drift can have severe consequences, potentially introducing security vulnerabilities, … [Read more...] about Using AI to Simplify Cloud Configuration Drift Detection
GenAI and RAG: Transforming security, fraud detection, and observability
GenAI is not magicI attended ElasticON recently where we spent the day with our NYC Elastic community, talking about the combined value of vector databases using retrieval augmented generation (RAG) to feed large language models (LLMs) for next-level generative AI (GenAI) results. Elastic’s CTO and Founder Shay Banon kicked off his keynote with an important message: GenAI is … [Read more...] about GenAI and RAG: Transforming security, fraud detection, and observability
Transforming fraud detection: AI and Elastic Security in financial services
Fraud in financial services is becoming more sophisticated, costing the industry billions annually and eroding customer trust. Recently, Deloitte published an article highlighting the risk AI brings in the form of fraudsters to the financial services industry: “Fake content has never been easier to create — or harder to catch. As threats grow, banks can invest in AI and other … [Read more...] about Transforming fraud detection: AI and Elastic Security in financial services
Researchers Explore Contrastive Learning for Malware Detection
CrowdStrike research shows that contrastive learning improves supervised machine learning results for PE (Portable Executable) malwareApplying self-supervised learning to PE files enhances the effectiveness of machine learning in cybersecurity, which is crucial to address the evolving threat landscapeCrowdStrike researchers engineered a novel loss function to optimize … [Read more...] about Researchers Explore Contrastive Learning for Malware Detection
Critical Evolution of Cloud Detection and Response
When Conventional Security Meets Modern Cloud Threats As organizations face these cross-domain attacks, the inability to connect cloud context with detection and alerting is reaching its breaking point. Most organizations begin their cloud security journey focused on visibility through CNAPP solutions. CNAPP identifies misconfigurations, excessive permissions and … [Read more...] about Critical Evolution of Cloud Detection and Response
Unify Security Posture and Protection for Faster Cloud Detection and Response
Adversaries are taking aim at cloud environments, as evidenced by the 75% increase in cloud intrusions in 2023. Organizations are under growing pressure to ensure they have measures in place to effectively detect, investigate and respond to cloud-focused attacks. Cloud detection and response (CDR) is uniquely challenging for SOC teams due to their limited visibility into … [Read more...] about Unify Security Posture and Protection for Faster Cloud Detection and Response
AI Could Generate 10,000 Malware Variants, Evading Detection in 88% of Case
Dec 23, 2024Ravie LakshmananMachine Learning / Threat Analysis Cybersecurity researchers have found that it's possible to use large language models (LLMs) to generate new variants of malicious JavaScript code at scale in a manner that can better evade detection. "Although LLMs struggle to create malware from scratch, criminals can easily use them to rewrite or obfuscate … [Read more...] about AI Could Generate 10,000 Malware Variants, Evading Detection in 88% of Case
Detection Logic for Pre-Deployment Malware Scanning
As organizations embrace DevOps practices and CI/CD pipelines to accelerate software delivery, their greater dependency on third-party components can introduce security risks. Because malware can infiltrate an environment during development, it’s important to check for it ahead of deployment. CrowdStrike Falcon® Cloud Security now applies its award-winning sensor … [Read more...] about Detection Logic for Pre-Deployment Malware Scanning
Cloud Data Logs: Heroes of Detection and Response
These logs provide deep visibility into the resource and service layers of cloud environments, enabling security teams to monitor for suspicious behavior, identify vulnerabilities and detect unauthorized actions. In the event of a breach, cloud logs are essential for incident response. They serve as the digital evidence needed to understand how an attack unfolded, which … [Read more...] about Cloud Data Logs: Heroes of Detection and Response