Mitre ATT&CK
PulseAugur coverage of Mitre ATT&CK — every cluster mentioning Mitre ATT&CK across labs, papers, and developer communities, ranked by signal.
13 day(s) with sentiment data
-
New framework TTP-R1 enhances cyber threat intelligence analysis
Researchers have developed TTP-R1, a novel two-stage framework designed to improve the extraction of attack techniques from cyber threat intelligence (CTI) text. This framework combines retrieval-augmented supervised fi…
-
AI Recommendation Poisoning: "Summarize with AI" buttons can corrupt LLM memory
Microsoft has identified a new threat called AI Recommendation Poisoning, where seemingly innocuous "Summarize with AI" buttons can permanently alter an LLM assistant's memory. By clicking a specially crafted link, user…
-
L1.9 prompt injection screening system rebrands, plans public corpus release
The L1.9 system, previously referred to as a firewall, is being renamed to prompt injection screening. This change reflects its function as a static pre-admission filter rather than a runtime enforcement layer. The deve…
-
Generative AI tool AegisShield democratizes cyber threat modeling
Researchers have developed AegisShield, a generative AI tool designed to automate and simplify cyber threat modeling, particularly for organizations with limited resources. The tool integrates the STRIDE framework and M…
-
MITRE ATLAS adds agentic attack techniques to AI threat catalog
MITRE has updated its Adversarial Threat Landscape for Artificial-Intelligence Systems (ATLAS) knowledge base to include specific techniques targeting autonomous AI agents. These new techniques, such as AI Supply Chain …
-
New AI system ThreatForest automates attack tree generation from code
Researchers have developed ThreatForest, a novel multi-agent system designed to automate threat modeling for cloud-native software development. This system analyzes code repositories to generate structured attack trees,…
-
New paper maps CVEs to MITRE ATT&CK using expert-labeled data
A new research paper details a method for mapping Common Vulnerabilities and Exposures (CVEs) to MITRE ATT&CK techniques using a classifier trained on 1,207 expert-labeled CVEs. The study found that using LLM-generated …
-
TRACE-CTI framework enhances audibility of cyber threat intelligence claims
A new framework called TRACE-CTI has been developed to improve the audibility and governance of cyber threat intelligence (CTI) claims extracted by automated systems. This framework preserves evidence, provenance, and v…
-
New Classifier Maps CVEs to MITRE ATT&CK Techniques, LLM Labeling Shows No Benefit
Researchers have developed a reproducible pipeline to map Common Vulnerabilities and Exposures (CVEs) to MITRE ATT&CK Enterprise techniques using free-text descriptions. Their custom-trained classifier, built on expert-…
-
Sakana AI launches Fugu-Cyber for cybersecurity tasks · 2 sources tracked
Sakana AI has launched Fugu-Cyber, an orchestration model specifically tuned for cybersecurity tasks. This model achieved 86.9% on the CyberGym benchmark and 72.1% on the CTI-REALM benchmark, positioning it as a competi…
-
Open-weight LLMs evaluated for autonomous vehicle threat intelligence generation
Researchers have developed a new dataset, CAV-STIXGen, to evaluate open-weight Large Language Models (LLMs) in generating structured threat information for autonomous vehicle vulnerabilities. The study assessed 11 LLMs,…
-
Developer builds 8-layer defense after trojan hits MCP marketplace
A developer recently discovered a Windows trojan, Trojan:Win64/Lazy.PGPK!MTB, within a skill package on their Model Context Protocol (MCP) marketplace. The malware was disguised as a legitimate GitHub repository through…
-
Coding agents trigger security alerts by mimicking attacker behavior · 1 source tracked
A recent report from Sophos highlights how coding agents like Claude Code, Cursor, and OpenAI Codex can inadvertently trigger security alerts on endpoints. These agents, while performing legitimate tasks such as automat…
-
NetForge RL: New Cyber Defense Simulation Environment Released
Researchers have introduced NetForge RL, a new multi-agent simulation environment designed for training reinforcement learning agents in cyber defense scenarios. This environment features procedurally generated networks…
-
New knowledge graph links software vulnerabilities to attack behaviors
Researchers have developed a new knowledge graph, CVE-TTP KG, designed to link software vulnerabilities with attacker tactics and techniques. This system aims to improve threat interpretation by connecting information f…
-
Anthropic report details AI misuse shift to agentic attacks
Anthropic has released a report detailing how malicious actors misuse AI models, particularly focusing on the shift from simple malware writing to more sophisticated agentic actions like lateral movement within networks…
-
GDM releases AI Control Roadmap with cybersecurity-inspired threat modeling
The GDM AI Control Roadmap (v0.1) has been released, outlining a plan for internal guardrails to detect and mitigate adversarial AI agent behavior. The roadmap draws inspiration from cybersecurity frameworks like MITRE …
-
New Agentra framework enhances enterprise intrusion response with multi-agent planning
A new research paper introduces Agentra, a multi-agent framework designed to automate and improve enterprise intrusion response. Agentra converts security alerts into structured incident response plans, leveraging frame…
-
New dataset combines system, network, and browser logs for cybersecurity
Researchers have developed a new multi-source cybersecurity dataset by combining system, network, and browser logs from Windows endpoints. This dataset, containing 870 sessions and approximately 2.3 million events, is l…
-
Open-source LLMs fall short on complex cyber threat intelligence classification
A new research paper evaluates the performance of seven open-source large language models (LLMs) on classifying complex cyber threat intelligence (CTI) reports. The study constructed a dataset of 2,076 human-annotated s…