Bài nghiên cứu mới giới thiệu mô hình Envelope Layer cho LLM Agent, đánh giá trọng lượng các thành phần và cách giải quyết rủi ro khi công cụ MCP và bộ nhớ lâu dài đồng thời tồn tại. Điều này giúp cải thiện an toàn và giảm rủi ro lây nhiễm nội dung độc hại trong AI Agent.
⏱ Độ trễ thật của tab này: trung bình 6.8 giờ sau khi CVE được công bố (n = 5458, sớm nhất 0.1 giờ). Số âm nghĩa là ta thấy trước ngày công bố.
arXiv:2611.19764v1 Announce Type: new Abstract: Autonomous coding agents (such as Antigravity, Claude Code, Cursor, and OpenHands) rely on third-party tools to inspect files, run commands, and call external APIs. Today's security auditors evaluate these tools almost entirely in …
arXiv:2611.20045v1 Announce Type: new Abstract: Deep research and agentic search agents autonomously retrieve content from the open web and generate reports with citations, and their outputs are often taken directly as research conclusions. This paper systematizes security resea…
OpenAI ra mắt Daybreak gồm Red Team (agent tự động pentest web app, phân tích mã nguồn, xác thực fix) và Defense Factory (xây patch + retest). Đáng chú ý vì là bước tiến cụ thể của AI agent vào pipeline security testing & vulnerability remediation của doanh nghiệp.
arXiv:2611.22583v1 Announce Type: new Abstract: Tool-using language-model agents must act on information from untrusted emails, webpages, and documents. This creates a central security challenge: malicious instructions and legitimate task directives can arrive through the same i…
arXiv:2611.22288v1 Announce Type: cross Abstract: Large-language-model agents now close the industrial predictive-maintenance loop, reading sensor windows, retrieving procedures and dispatching work orders. Two attack channels meet there: injected text in a retrieved document ca…
A malicious widget can bypass the security filter that Telegram added to its Android app last week and run code outside it, according to NebuSec, a security company. Widgets are small interactive blocks that bots send inside chat messages, and their JavaScript runs on the user's…
🎯 Kẻ tấn công làm gì (từng bước)
- Tìm đúng đường dẫn/API có xử lý dữ liệu người dùng gửi lên
- Gửi payload đã chế để máy chủ thực thi lệnh/mã
- Dùng quyền của tiến trình web để đọc file, cài backdoor, mở kênh ra ngoài
🔍 Dấu hiệu bị tấn công (để kiểm tra log)
- Tiến trình web (php-fpm/nginx/java) sinh tiến trình con bất thường: sh, bash, curl, wget
- Kết nối ra ngoài từ máy chủ tới IP lạ đúng lúc có request lạ
- File mới xuất hiện trong thư mục tải lên/tạm
🛡 Phòng ngay (làm được hôm nay)
- Vá ngay bản có lỗi (ưu tiên số 1)
- Chặn tạm đường dẫn/API bị lỗi ở WAF/reverse proxy nếu chưa vá được
- Chạy tiến trình web bằng user quyền thấp, tắt các hàm shell (disable_functions) nếu dùng PHP
- Bật giám sát tiến trình con của web server để phát hiện lệnh lạ
arXiv:2611.23236v1 Announce Type: cross Abstract: LLM-enhanced recommender systems encode item side information into semantic embeddings, but these embeddings may leak sensitive textual attributes through inversion attacks. Existing input-side defenses rely on trusted collection…
arXiv:2611.18027v1 Announce Type: new Abstract: Web caches reduce origin load and latency, but automated traffic can either warm them or evict objects shoppers reuse, lowering the cache hit ratio (CHR). Existing bot defenses generally infer automation from User-Agent (UA) string…
Vals, a San Francisco-based AI evaluation startup, tested whether leading AI models could be manipulated into participating in autonomous warfare. They usually compiled.
arXiv:2611.24662v1 Announce Type: new Abstract: AI-driven attacks learn the models of cyber-physical systems (CPS) by passively observing controller and plant signals, then use this knowledge to act covertly. Because covertness depends on the accuracy of the identified model, de…
arXiv:2611.16122v1 Announce Type: new Abstract: AI agents increasingly access enterprise data, invoke powerful tools, and act with limited human oversight. Existing security defenses primarily detect prompt injection, harmful outputs, or suspicious behavior, but do not determine…
arXiv:2611.20291v1 Announce Type: new Abstract: LLM serving engines cache prompt prefixes; a cache hit skips computation and returns faster, so a shared cache reveals whether others have sent a given text. Per-account salting closes this channel but forgoes cross-tenant reuse an…
Cybercriminals using agentic pen-testing tools is a proverbial genie out of the bottle at this point, so enterprises need to get a defense plan in place right away.
Igor Andreevich D’iu, A Russian national recruited through Telegram, has been charged in the United States with acting as a Moscow agent in a surveillance campaign targeting employees of a defense technology company, understood to be Palantir Technologies.
Attackers used Cairn once again in autonomous attacks that targeted advanced manufacturing and defense sectors, as well as civil and human rights organizations.
The Anduril co-founder says outdated technology is leaving systems vulnerable to a new wave of AI-enabled attacks.
Daybreak Red Team pentests web applications with AI agents, while Defense Factory investigates security findings and prepares patches. The post OpenAI Unveils Daybreak Red Team, Defense Factory Security Tools appeared first on SecurityWeek .
arXiv:2611.15013v1 Announce Type: new Abstract: Machine unlearning lets the owner of large language models (LLMs) remove specific entities from a trained model without retraining from scratch. Existing machine unlearning methods focus on removing the requested entities while mai…
arXiv:2611.21832v1 Announce Type: new Abstract: As AI systems are increasingly used to automate AI R&D, misaligned systems could sabotage research intended to improve model safety and alignment. To stress-test defenses against such sabotage, we use a human expert as a proxy sabo…
The problem disrupted the operations of NSANet, a classified internal network used by employees of the agency. The post Classified Communications Network at the National Security Agency Goes Down, Defense Official Says appeared first on SecurityWeek .
The emerging technology will soon break the "lock and key" encryption currently protecting sensitive data. A new GAO report shows agencies are behind on building their defenses.
Pinning GitHub Actions to an exact version of their code, known as SHA pinning, has become the go-to defense against software supply chain attacks. Attackers have already found a way around it. In March 2026, the hackers who hijacked Trivy, a widely used open-source security sca…
🎯 Kẻ tấn công làm gì (từng bước)
- Tìm đúng đường dẫn/API có xử lý dữ liệu người dùng gửi lên
- Gửi payload đã chế để máy chủ thực thi lệnh/mã
- Dùng quyền của tiến trình web để đọc file, cài backdoor, mở kênh ra ngoài
🔍 Dấu hiệu bị tấn công (để kiểm tra log)
- Tiến trình web (php-fpm/nginx/java) sinh tiến trình con bất thường: sh, bash, curl, wget
- Kết nối ra ngoài từ máy chủ tới IP lạ đúng lúc có request lạ
- File mới xuất hiện trong thư mục tải lên/tạm
🛡 Phòng ngay (làm được hôm nay)
- Vá ngay bản có lỗi (ưu tiên số 1)
- Chặn tạm đường dẫn/API bị lỗi ở WAF/reverse proxy nếu chưa vá được
- Chạy tiến trình web bằng user quyền thấp, tắt các hàm shell (disable_functions) nếu dùng PHP
- Bật giám sát tiến trình con của web server để phát hiện lệnh lạ
NATO’s Communications and Information Agency (NCIA) gives NATO’s core institutions access to OpenAI’s cybersecurity models to help protect Allied networks. NATO has used up to 300 billion raw input and output tokens a month. Access comes through Daybreak, OpenAI’s program for vet…
arXiv:2611.19864v1 Announce Type: new Abstract: Website fingerprinting (WF) attacks let a passive adversary infer which site a Tor user visits from the encrypted cell sequence. On undefended traffic, hand-crafted-feature attacks are already near-ceiling, so deep learning offers …
arXiv:2611.20594v1 Announce Type: new Abstract: An enterprise breach causes most of its damage after the perimeter is crossed. Once an attacker gains a foothold, they can move laterally to other hosts, take over the trusted services of the network, and encrypt the backups needed…
How OpenAI is working with NATO and allied cyber agencies to help trusted defenders find vulnerabilities, test fixes and protect essential systems.
Tor 0.4.9.17 patches several security issues, including one related to how built-in DoS protections deal with IPv6. The post Tor Updates DoS Defenses Again as Relay Attacks Continue appeared first on SecurityWeek .
🎯 Kẻ tấn công làm gì (từng bước)
- Gửi dữ liệu gây crash hoặc tiêu tốn tài nguyên
- Dịch vụ ngừng phục vụ
🔍 Dấu hiệu bị tấn công (để kiểm tra log)
- Dịch vụ restart liên tục, RAM/CPU tăng vọt
- Nhiều request cùng một tham số từ ít IP
🛡 Phòng ngay (làm được hôm nay)
- Vá/cập nhật
- Giới hạn tần suất + giới hạn kích thước request
- Có giám sát uptime và tự khởi động lại
arXiv:2611.17761v1 Announce Type: new Abstract: Many cyber-physical systems (CPS) are multi-mode: they switch between operating modes with different tasksets and timing parameters in response to physical events or user activity. We show that these mode changes create a timing si…
arXiv:2611.17907v1 Announce Type: new Abstract: We consider a physical layer authentication scenario in which a legitimate user (Alice) communicates with an authenticator (Bob) through a reconfigurable intelligent surface (RIS), controlled by Bob. An attacker (Trudy) aims to imp…
German defense technology company Quantum Systems confirmed that it detected an incident in late October that was part of an active cyberattack by "hostile actors." [...]
arXiv:2611.16557v1 Announce Type: new Abstract: We consider a physical layer authentication scenario in which a legitimate user (Alice) communicates with an authen- ticator (Bob) through a reconfigurable intelligent surface (RIS), controlled by Bob. An attacker (Trudy) aims to i…
The perpetrators targeted job seekers in India, mostly, but some of the effort could be more broadly targeting defense contractors and military personnel. The post Pakistani hackers are spying on defense workers with fake job sites appeared first on CyberScoop .
German defense technology company Quantum Systems confirmed that it detected an incident in late October that was part of an active cyberattack by "hostile actors." [...]
Despite longstanding concerns about the Kremlin’s access to the app’s data, NATO on Thursday issued guidance suggesting attendees of its flagship defense industry conference use the Russian-owned Yandex Go to reach the venue.
CyberScoop is exclusively reporting on the letter to the Hill on the controversial snooping authorities under the Foreign Intelligence Surveillance Act. The post Don’t put Section 702 spying law renewal in defense bill, groups urge Hill leaders appeared first on CyberScoop .