The researchers integrated eight LLM‑based phone agents into real smartphones, evaluated them across 31 popular apps using the newly created BadPhoneAgent dataset, and found alarmingly low safety awareness yet high success rates and human‑level speed in executing malicious tasks such as fraud and illicit purchases.
WallBreaker is an open-source CLI that automates LLM red-team testing by iteratively mutating attack payloads, offering a library of research-grade techniques, a 59-to-222 transformation engine, multimodal image attacks, a HarmBench-based judge, and performance optimizations that cut token costs by 20% and boost success rates by about 30%.
OpenAI's GPT-5.6 arrives with three tiered models (Sol, Terra, Luna), new max and ultra reasoning modes, benchmark breakthroughs in programming, biology, and security, extensive multi‑layer safety guards, a steep pricing structure, and a tightly controlled preview rollout.
The article explains how Metasploit’s new auxiliary modules—dhcp_exhaustion/exhaust and dns_mitm/dns_mitm—can be combined to exhaust a DHCP server’s address pool, impersonate it, and redirect DNS queries to a malicious server, enabling attackers to gain network control while outlining defensive measures such as DHCP snooping and ARP inspection.
The Innora.ai research report dissects 46 high‑impact CVEs spanning OS kernels, multimedia libraries, enterprise middleware, AI inference servers and mobile apps, revealing how an AI‑driven automated red‑team framework (DialTree‑RPO) uncovers and validates these flaws at unprecedented speed and scale.
The article explains how production‑grade AI agents require a full‑stack security framework—covering input sanitization, runtime policy enforcement, output verification, and audit—to mitigate ten OWASP attack surfaces such as prompt injection, tool misuse, memory poisoning, and cascading failures, with practical defense layers and red‑team testing guidance.
Anthropic’s decision to withhold the powerful Claude Mythos model sparked a joint industry effort called Project Glasswing, revealing how AI can dramatically accelerate vulnerability discovery and prompting security professionals to rethink their roles, adopt AI tools, and evolve their skill sets.
The article examines how easily accessible, unfiltered large language models enable even novice attackers to create sophisticated, personalized phishing campaigns and rapid reconnaissance, while defenders scramble to adopt small, locally‑run AI models for detection, UEBA, and reverse‑engineering of AI‑generated malware.
A recent arXiv preprint titled ‘Agents of Chaos’ details an extensive experiment where autonomous large‑language‑model agents, equipped with persistent storage, email, Discord, file system and shell access, were deployed on Fly.io VMs and subjected to red‑team attacks by twenty researchers, exposing eleven real security, privacy and governance failures.
The article analyzes Surge AI’s $1.2 billion revenue, bootstrapped model, elite 100 k‑labeler network, three‑layer architecture, RLHF, AdvancedIF/RIFL benchmarks, red‑team testing, RL environments, and evaluates its competitive moat and future strategic paths.
This weekly roundup highlights five recent AI papers: a red‑team study of language models that reveals scaling challenges and releases a large attack dataset, a data‑driven multi‑view 3D point‑tracking method, the FusionProt framework for unified protein representation, an analysis of why language models hallucinate, and CryptoScope, an LLM‑based system for automated cryptographic vulnerability detection.
