Tag: injection
-
GitLab Vulnerability ‘Highlights the Double-Edged Nature of AI Assistants’
A remote prompt injection flaw in GitLab Duo allowed attackers to steal private source code and inject malicious HTML. GitLab has since patched the issue. First seen on techrepublic.com Jump to article: www.techrepublic.com/article/news-gitlab-duo-vulnerability-hidden-prompts/
-
GitLab ‘Vulnerability Highlights the Double-Edged Nature of AI Assistants’
A remote prompt injection flaw in GitLab Duo allowed attackers to steal private source code and inject malicious HTML. GitLab has since patched the issue. First seen on techrepublic.com Jump to article: www.techrepublic.com/article/news-gitlab-duo-vulnerability-hidden-prompts/
-
Meteobridge Web Interface Vulnerability Let Attackers Inject Commands Remotely
ONEKEY Research Lab has uncovered a severe command injection vulnerability in the MeteoBridge firmware, a compact device designed to connect personal weather stations to public weather networks like Weather Underground. This flaw, identified through ONEKEY’s recently introduced bash static code analysis on their platform, affects versions 6.1 and below of the MeteoBridge firmware, enabling remote,…
-
How Hunters International Used the Browser to Breach Enterprises”Š”, “ŠAnd Why They Didn’t See It”¦
How Hunters International Used the Browser to Breach Enterprises”Š”, “ŠAnd Why They Didn’t See It Coming At RSAC 2025, Cato Networks delivered a presentation that SOC teams and CISOs will want to pay attention to: “Suspicious Minds”Š”, “ŠHunting Threats That Don’t Trigger Security Alerts.” The session showcased ransomware campaigns that bypassed traditional detection. In some cases,…
-
GitLab Duo Vulnerability Exploited to Inject Malicious Links and Steal Source Code
A security vulnerability was recently discovered in GitLab Duo, the AI-powered coding assistant integrated into GitLab and based on Anthropic’s Claude models. Security researchers from Legit Security revealed that attackers could exploit an indirect prompt injection flaw to exfiltrate private source code, manipulate AI-generated code suggestions, and even leak confidential zero-day vulnerabilities”, all through seemingly…
-
GitLab Duo Vulnerability Enabled Attackers to Hijack AI Responses with Hidden Prompts
Cybersecurity researchers have discovered an indirect prompt injection flaw in GitLab’s artificial intelligence (AI) assistant Duo that could have allowed attackers to steal source code and inject untrusted HTML into its responses, which could then be used to direct victims to malicious websites.GitLab Duo is an artificial intelligence (AI)-powered coding assistant that enables users to…
-
GitLab’s AI Assistant Opened Devs to Code Theft
Even after a fix was issued, lingering prompt injection risks in GitLab’s AI assistant might allow attackers to indirectly deliver developers malware, dirty links, and more. First seen on darkreading.com Jump to article: www.darkreading.com/application-security/gitlab-ai-assistant-opened-devs-to-code-theft
-
GitHub’s AI Assistant Opened Devs to Code Theft
Even after a fix was issued, lingering prompt injection risks in GitLab’s AI assistant might allow attackers to indirectly deliver developers malware, dirty links, and more. First seen on darkreading.com Jump to article: www.darkreading.com/application-security/github-ai-assistant-opened-devs-to-code-theft
-
New Process Injection Technique Evades EDR by Injecting Malicious Code into Windows Processes
Researchers revealed this method exploits shared memory regions and thread context manipulation to execute malicious payloads without triggering standard detection heuristics. Novel process injection technique leveraging execution-only primitives has demonstrated the ability to bypass leading Endpoint Detection and Response (EDR) systems by avoiding traditional memory allocation and modification patterns. Modern EDR solutions typically monitor for…
-
Hackers Target Mobile Users Using PWA JavaScript to Bypass Browser Security
A sophisticated new injection campaign has been uncovered, targeting mobile users through malicious third-party JavaScript to deliver a Chinese adult-content Progressive Web App (PWA) scam. This attack, which redirects users to sites like hxxps://xjdm166[.]com, leverages the unique capabilities of PWAs to retain users longer and evade traditional browser security mechanisms. Unlike typical phishing attempts, this…
-
Researchers Expose PWA JavaScript Attack That Redirects Users to Adult Scam Apps
Cybersecurity researchers have discovered a new campaign that employs malicious JavaScript injections to redirect site visitors on mobile devices to a Chinese adult-content Progressive Web App (PWA) scam.”While the payload itself is nothing new (yet another adult gambling scam), the delivery method stands out,” c/side researcher Himanshu Anand said in a Tuesday analysis.”The malicious landing…
-
8 KI-Sicherheitsrisiken, die Unternehmen übersehen
Tags: access, ai, api, application-security, authentication, cisco, ciso, compliance, cyber, cyberattack, cybersecurity, data, data-breach, framework, governance, hacker, injection, LLM, RedTeam, risk, risk-management, security-incident, software, threat, tool, vulnerabilityIn ihrem Wettlauf um Produktivitätssteigerungen durch generative KI übersehen die meisten Unternehmen die damit verbundenen Sicherheitsrisiken.Laut einer Studie des Weltwirtschaftsforums, die in Zusammenarbeit mit Accenture durchgeführt wurde, versäumen es 63 Prozent der Unternehmen, die Sicherheit von KI-Tools vor deren Einsatz zu überprüfen. Dadurch gehen sie eine Reihe von Risiken für ihr Unternehmen ein.Dies gilt sowohl…
-
Confluence Servers Under Attack: Hackers Leverage Vulnerability for RDP Access and Remote Code Execution
Tags: access, attack, breach, cve, cyber, data-breach, exploit, flaw, hacker, injection, network, remote-code-execution, threat, vulnerabilityThreat actors exploited a known vulnerability, CVE-2023-22527, a template injection flaw in Atlassian Confluence servers exposed to the internet. This exploit facilitated remote code execution (RCE), enabling attackers to gain initial access and establish a foothold within targeted networks. The breach, first detected through network traffic from IP address 45.227.254.124 running a >>whoami
-
Researchers Expose New Intel CPU Flaws Enabling Memory Leaks and Spectre v2 Attacks
Researchers at ETH Zürich have discovered yet another security flaw that they say impacts all modern Intel CPUs and causes them to leak sensitive data from memory, showing that the vulnerability known as Spectre continues to haunt computer systems after more than seven years.The vulnerability, referred to as Branch Privilege Injection (BPI), “can be exploited…
-
12 AI terms you (and your flirty chatbot) should know by now
1. Artificial general intelligence (AGI) The ultimate manifestation of AI has already played a featured role in dozens of apocalyptic movies. AGI is the point at which machines become capable of original thought and either a) save us from our worst impulses or b) decide they’ve had enough of us puny humans. While some AI…
-
F5 BIG-IP Vulnerability Allows Remote Command Execution
Critical security vulnerability in F5 BIG-IP systems has been discovered that allows authenticated administrators to execute arbitrary system commands, effectively bypassing security boundaries. Identified as CVE-2025-31644, the command injection flaw affects multiple versions of BIG-IP running in Appliance mode. Security researcher Matei >>Mal
-
Mitel SIP Phone Flaws Allow Attackers to Inject Malicious Commands
A pair of vulnerabilities in Mitel’s 6800 Series, 6900 Series, and 6900w Series SIP Phones-including the 6970 Conference Unit-could enable attackers to execute arbitrary commands or upload malicious files to compromised devices, posing significant risks to enterprise communication systems. The flaws, disclosed in Mitel’s Product Security Advisory MISA-2025-0004, include a critical-severity command injection bug (CVE-2025-47188)…
-
Indirect Prompt Injection Exploits LLMs’ Lack of Informational Context
A new wave of cyber threats targeting large language models (LLMs) has emerged, exploiting their inherent inability to differentiate between informational content and actionable instructions. Termed >>indirect prompt injection attacks,
-
Hackers Exploit Host Header Injection to Breach Web Applications
Cybersecurity researchers have reported a significant rise in web breaches triggered by a lesser-known technique: Host Header Injection. This sophisticated attack vector has enabled hackers to compromise numerous web applications, steal sensitive information, and manipulate website operations-raising alarm bells among businesses and security experts alike. At the core of this attack is the manipulation of…
-
LLM02: Sensitive Information Disclosure FireTail Blog
May 08, 2025 – Lina Romero – In 2025, AI security is a relevant issue. With the landscape changing so rapidly and new risks emerging every day, it is difficult for developers and security teams to stay on top of AI security. The OWASP Top 10 Risks for LLM attempts to break down the most prevalent…
-
Stealth Is the Strategy: Rethinking Infrastructure Defense
Tags: access, ai, attack, breach, cisco, cloud, cybersecurity, data, defense, edr, endpoint, espionage, exploit, finance, firewall, gartner, google, group, infrastructure, injection, ivanti, malicious, monitoring, network, resilience, risk, strategy, technology, threat, tool, vpn, vulnerability, zero-day, zero-trust -
Hackers Exploit 21 Apps to Take Full Control of E-Commerce Servers
Tags: attack, backdoor, control, cyber, cybersecurity, exploit, hacker, injection, malicious, software, supply-chain, vulnerabilityCybersecurity firm Sansec has uncovered a sophisticated supply chain attack that has compromised 21 popular e-commerce applications, granting hackers full control over hundreds of online stores. This malicious campaign, which began with the injection of backdoors as early as six years ago, was activated this week, exposing vulnerabilities in software from vendors such as Tigren,…
-
Top cybersecurity products showcased at RSA 2025
Tags: access, ai, attack, automation, awareness, breach, cisco, compliance, control, crowdstrike, cyber, cybersecurity, data, defense, detection, edr, email, firewall, fortinet, framework, identity, incident response, infrastructure, injection, intelligence, login, malicious, open-source, phishing, risk, siem, soc, threat, tool, training, update, vulnerability, zero-trustCisco: Foundational AI Security Model: Cisco introduced its Foundation AI Security Model, an open-source framework designed to standardize safety protocols across AI models and applications. This initiative aims to address the growing concerns around AI security and ensure Safer AI deployments. Cisco also unveiled new agentic AI features in its XDR and Splunk platforms, along…
-
New MCP-Based Attack Techniques and Their Application in Building Advanced Security Tools
MCP, developed by Anthropic, allows Large Language Models (LLMs) to interface seamlessly with external tools, enabling the creation of agentic AI systems that can autonomously perform complex tasks. As organizations increasingly integrate MCP, new attack techniques have emerged, highlighting the importance of robust security controls and innovative defensive strategies. MCP Tool Manipulation and Prompt Injection…