Machine learning supports malware detection and threat analysis, but attackers can also exploit biased data, model weaknesses, or poisoned training.
Search across headline titles and summaries.
Background for this topic.
Machine learning is a way to build software that learns patterns from data and uses them to classify, predict, or make decisions, rather than relying solely on hand-written rules. Models may support malware and phishing detection, user- or entity-behavior analysis, vulnerability prioritization, and automated security triage. Their outputs are probabilistic, so unusual activity can be missed or incorrectly flagged; changing normal behavior can also cause model drift and reduce accuracy.
Security teams must protect both the model and its training data. An attacker may manipulate training examples (data poisoning), craft inputs designed to evade detection (adversarial examples), or extract sensitive information from a model or its data. Controls include trusted data provenance, access restrictions, testing against realistic evasive inputs, monitoring for drift, and human review of high-impact decisions. Where models process personal, proprietary, or security telemetry, collection, retention, and reuse require appropriate privacy and governance controls.
Weekly headline count for the current query.
A flaw in the Google Cloud Vertex AI SDK for Python let an attacker with no access to a victim's project hijack the victim's machine learning model upload and run code inside Google's serving infrastructure
Proton uses machine learning models to detect abuse of its services – especially email addresses used by cybercriminals
A new family of Android click-fraud trojans leverages TensorFlow machine learning models to automatically detect and interact with specific advertisement elements. [...]
Artificial intelligence (AI) is making its way into security operations quickly, but many practitioners are still struggling to turn early experimentation into consistent operational value. This is because SOCs are adopting AI without an intentional approach to operational integration. Some teams treat it as a shortcut for broken processes. Others attempt to apply machine learning to problems
AI in OT May Trigger Cascading Infrastructure FailuresThe U.S. cyber defense agency warned that machine learning and large language model deployments can introduce new attack surfaces across critical infrastructure sectors in a document setting out principles for safely integrating AI into operational technology.
Ethicist Harry Farmer on Data Privacy, Predictive Analytics and Fairness IssuesArtificial intelligence, particularly machine learning, is transforming genomics by enabling powerful predictions about health and human traits from DNA data. But this convergence of technologies raises major red flags related to data privacy and security, said senior researcher Harry Farmer.
How to avoid your business being felled by an AI-powered ransomware attack that costs less than a laptop. Passwork KNP Logistics Group, a British transport company from Northamptonshire that’s been around longer than the mass-produced lightbulb, collapsed after a devastating security breach that left more than 700 employees jobless. The 158-year-old firm fell victim to a ransomware attack.…
XML-Based Messaging Tech Extends Fraud Detection Into Wider Bank Use CasesACI's signals network intelligence harnesses neural networks and federated machine learning to spot fraud in real time without banks sharing data. Beyond fraud detection, its insights can drive business growth from other business units, and ACI aims to accelerate adoption by making it open source.
Google’s Gemini-powered tools tripped up by image-scaling prompt injection Security researchers with Trail of Bits have found that Google Gemini CLI and other production AI systems can be deceived by image scaling attacks, a well-known adversarial challenge for machine learning systems.…
AI Helps Delta Shift 20% of Ticket Pricing to Real-Time Automation by 2025Delta Air Lines is revolutionizing ticket pricing with AI, aiming to automate 20% of fares by 2025. Partnering with Fetcherr, the airline uses real-time data and machine learning for personalized pricing, raising revenue potential and privacy concerns.
WeTransfer added the magic words "machine learning" to its ToS and users reacted predictably Analysis WeTransfer this week denied claims it uses files uploaded to its ubiquitous cloud storage service to train AI, and rolled back changes it had introduced to its Terms of Service after they deeply upset users. The topic? Granting licensing permissions for an as-yet-unreleased LLM product.…
Processing and storage for Gemini 2.5 Flash to stay in Blighty Google Cloud is attempting to ease concerns about where AI data is stored by offering organizations the option to keep Gemini 2.5 Flash machine learning processing entirely within the UK.…
A critical Azure Machine Learning flaw allows privilege escalation, risking subscription compromise
Emerging artificial intelligence and machine learning technologies being applied in the health and wellness space that are not necessarily covered by HIPAA but instead fall under a variety of tough new state privacy laws that are being enacted, said attorney Lily Li of Metaverse Law.
A novel attack exploited machine learning models on PyPI, using zipped Pickle files to deliver infostealer malware
Emerging artificial intelligence and machine learning technologies being applied in the health and wellness space that are not necessarily covered by HIPAA but instead fall under a variety of tough new state privacy laws that are being enacted, said attorney Lily Li of Metaverse Law.
Machine Learning, Generative AI Bolster Continuous User AuthenticationFinancial institutions can use AI-fueled behavioral biometrics for real-time identity assurance. By continuously profiling how users interact with devices, firms can shift from one-time authentication to real-time identity assurance, turning every click, pause and keystroke into a frontline defense.
Politecnico di Milano's Zanero on Evolving Malware Detection and Hardware SecurityMachine learning excels at identifying repetitive patterns and anomalies, but human insight remains vital for understanding the broader context of cyberattacks - especially in cyber-physical ecosystems, said Stefano Zanero, professor at Politecnico di Milano.
Microsoft has resolved an issue with a machine learning model that mistakenly flagged emails from Gmail accounts as spam in Exchange Online. [...]
Microsoft says it mitigated a known issue in one of its machine learning (ML) models that mistakenly flagged Adobe emails in Exchange Online as spam. [...]