Tag: injection
-
Toolkit Hidden Inside Oracle Database Evades Endpoint Tools
Attackers used SQL injection to compile a post-exploitation toolkit inside an Oracle database First seen on infosecurity-magazine.com Jump to article: www.infosecurity-magazine.com/news/khunt-toolkit-oracle-database-sql/
-
AI Recommendation Poisoning: How “Ask AI” Buttons Silently Alter LLM Memory
A new class of prompt injection is spreading across commercial websites. It requires no malware, no stolen credentials, and no zero-day exploit. It abuses a standard feature built into almost every major AI assistant: pre-filled deep links.We observed production websites embedding hidden prompt injection payloads inside “Ask AI” buttons on marketing and competitor comparison pages.…
-
KHunt Toolkit Turns Oracle SQL Injection Into SYSTEM-Level RCE and Credential Theft
Tags: credentials, cyber, data, infrastructure, injection, oracle, rce, remote-code-execution, sql, theft, threatKHunt shows how a “routine” SQL injection against an Oracle”‘backed web app can be weaponized into SYSTEM”‘level remote code execution and credential theft by compiling a full post”‘exploitation toolkit directly inside the database engine. This incident materially shifts the Oracle threat model: the database itself becomes attacker infrastructure, not just a data store. Subsequent triage…
-
Attackers Compile khunt Inside Oracle to Turn SQL Injection Into Windows SYSTEM Access
Attackers broke into an organization’s Oracle database through a SQL injection flaw in a public-facing web application, then installed a post-exploitation toolkit without writing an executable to disk. They fed Java source code to the database, let Oracle compile it into stored schema objects, and ran commands from inside the database engine.Huntress, which tracks the…
-
Cisco Patches 7 IOS XE Vulnerability Classes, Including Critical Command Injection Flaws
Cisco has released security-hardening updates for IOS XE Software that address seven classes of vulnerabilities, including a critical command, operating system, and argument injection category identified as CVE-2026-20272. The advisory, released on August 5, has an overall CVSS score of 3.1 and 9.8 and provides no workarounds, meaning that upgrading is the only recommended solution.…
-
AI Attacks Are Evolving: TryHackMe Demo Explores Prompt Injection
TryHackMe demo: Explore how prompt injection is reshaping AI security risks. First seen on esecurityplanet.com Jump to article: www.esecurityplanet.com/threats/ai-attacks-are-evolving-tryhackme-demo-explores-prompt-injection/
-
No Perfect Fix for AI Browser Prompt Injection Flaws
AI browsers from top vendors remain vulnerable to prompt injection attacks despite multiple security guardrails, according to new research. First seen on darkreading.com Jump to article: www.darkreading.com/application-security/no-perfect-fix-ai-browser-prompt-injection-flaws
-
Oracle SQL Injection Attack Enables Remote Code Execution
A Huntress investigation reveals how attackers used SQL injection to achieve RCE and steal credentials. First seen on esecurityplanet.com Jump to article: www.esecurityplanet.com/threats/oracle-sql-injection-attack-enables-remote-code-execution/
-
Hackers run khunt post-exploitation toolkit from Oracle database
Hackers exploited a SQL injection vulnerability to install a post-exploitation toolkit directly inside an Oracle database that was used to breach a corporate network. First seen on bleepingcomputer.com Jump to article: www.bleepingcomputer.com/news/security/hackers-run-khunt-post-exploitation-toolkit-from-oracle-database/
-
OpenAI’s GPT-5.6 Tests Show Prompt-Injection Gains and Agent Risks
OpenAI’s latest GPT-5.6 safety results show low failure rates for direct prompt injection but higher success rates when attacks arrive through tools and external content. The post OpenAI’s GPT-5.6 Tests Show Prompt-Injection Gains and Agent Risks appeared first on TechRepublic. First seen on techrepublic.com Jump to article: www.techrepublic.com/article/news-gpt-5-6-prompt-injection/
-
Menlo Security Extends MARS to Protect AI Assistants and Coding Agents
Menlo Security has expanded Menlo Agent Runtime Security, or MARS, with controls aimed at protecting AI assistants and coding agents from prompt injection, malware and data loss as they browse the web, use applications and process files. The company is highlighting the update at Black Hat USA 2026. MARS is a cloud-based capability in Menlo’s..…
-
Aembit Adds Workload Identity Federation Support for the Claude API
5 min readToday, we’re announcing the Aembit Claude Workload Identity Federation Credential Provider, the latest addition to Aembit’s growing Claude support. Aembit already covers a lot of ground with Claude: workloads can authenticate to the Claude API using static API key injection, Claude Web and the Claude App are supported as Client Workloads, direct API…
-
Menlo Security Extends Platform Reach to AI Agents
Menlo Security today at the Black Hat USA conference extended its cloud platform to secure artificial intelligence (AI) agents at runtime by sanitizing the web pages and files they read to neutralize prompt injection attacks and block data exfiltration. Company CEO Bill Robbins said the Menlo Agent Runtime Security (MARS) platform created to isolate browsers..…
-
Hackers Smuggle Post-Exploitation Toolkit Into Oracle Database Via Classic SQL Injection Flaw
A SQL injection vulnerability that many organisations might consider a decades-old, well-understood threat has been used as the entry point for a far more sophisticated attack, after threat actors were caught planting a custom-built, database-resident toolkit inside an Oracle database. Security firm Huntress said it was alerted to suspicious activity on an endpoint hosting an…
-
Prompt Injection Remains Biggest LLM Risk, Despite Limited Incidents
Prompt injection remains the most dangerous security threat to LLMs, according to OWASP’s latest Top 10 LLM Applications list First seen on infosecurity-magazine.com Jump to article: www.infosecurity-magazine.com/news/prompt-injection-llm-risk/
-
CISA Flags Langflow RCE, Tomcat, and N-central Flaws as Actively Exploited
Tags: cisa, cve, cybersecurity, exploit, flaw, infrastructure, injection, kev, rce, remote-code-execution, vulnerabilityThe U.S. Cybersecurity and Infrastructure Security Agency (CISA), on August 5, 2026, added three flaws to its Known Exploited Vulnerabilities (KEV) catalog, citing evidence of active exploitation in the wild.The list of vulnerabilities is as follows – CVE-2026-9198 (CVSS score: 9.8) – A code injection vulnerability in Langflow that allows unauthenticated attackers to achieve full…
-
30 days with Claude Mythos Preview: How Tenable adapted our security program, and why yours is next
Tags: ai, api, attack, business, control, cybersecurity, data, data-breach, endpoint, exploit, flaw, injection, LLM, remote-code-execution, risk, service, threat, tool, update, vulnerabilityTenable spent 30 days running frontier AI models against our own code. It didn’t just find bugs, it proved they’re real, with reproducible exploits. That fundamentally changes code security from ranking potential code defects to a much higher signal focused on the findings that matter. Read on to learn how it reshaped our security team’s…
-
Hackers Exploit Critical Arista VeloCloud Flaw to Execute OS Commands
Arista Networks has issued a warning about attackers actively exploiting CVE-2026-16812, a critical unauthenticated OS command injection vulnerability in on-premises VeloCloud Orchestrator (VCO) deployments. This flaw carries a CVSS v3.1 and v4.0 severity score of 10.0, allowing a remote attacker with network access to the VCO web interface to access privileged internal functions and potentially…
-
Anthropic’s Opus 5 Is Better at Resisting Prompt Injection
Tags: injectionThe chart is interesting. On the IPI benchmark, Opus 5 improved over Opus 4.8, reducing the probability of an attacker succeeding within 15 attempts from 5.5% to 2.0%, and from 0.5% to 0.2% on 1 attempt. It also improved on Sonnet 5 (5.9% at k=15) and Mythos 5 (2.6%), making it the most robust model…
-
PHP Patches 3 Security Flaws Enabling SQL Injection, Memory Corruption and DoS Attacks
PHP maintainers have released security updates to address three vulnerabilities affecting the PostgreSQL, BCMath, and Phar extensions. These vulnerabilities could potentially lead to SQL injection attacks, out-of-bounds memory writes, and denial-of-service attacks in vulnerable applications. The issues impact several actively maintained PHP release branches and have been resolved in versions PHP 8.2.338.2, 8.3.338.3, 8.4.248.4, and…
-
Attackers Exploit Arista VeloCloud Orchestrator Command Injection Flaw
A maximum-severity security flaw impacting on-premises versions of Arista VeloCloud Orchestrator (VCO) has come under active exploitation in the wild.The vulnerability, tracked as CVE-2026-16812 (CVSS score: 10.0), is a case of operating system command injection that could pave the way for arbitrary code execution.”VeloCloud Orchestrator (VCO) on-prem has a security issue where this issue First…
-
Arista patches VeloCloud Orchestrator zero-day exploited in attacks
Arista has patched a maximum-severity command injection vulnerability in on-premises VeloCloud Orchestrator deployments that is being actively exploited in attacks. First seen on bleepingcomputer.com Jump to article: www.bleepingcomputer.com/news/security/arista-patches-velocloud-orchestrator-zero-day-exploited-in-attacks/
-
Apache Syncope Flaws Let Users Gain Admin Roles and Execute Remote Code
Tags: apache, authentication, cve, cyber, flaw, injection, remote-code-execution, service, sql, vulnerabilityApache Syncope has released versions 4.1.24.1, 4.1.24.1.2, and 4.0.74.0.7 to address six security vulnerabilities affecting the 4.1, 4.0, and 3.0 release branches. These vulnerabilities include a self-service privilege escalation bug, multiple post-authentication remote code execution (RCE) pathways, authenticated server-side request forgery (SSRF), and SQL injection issues. Apache Syncope Flaws CVE-2026-62183 affects deployments that utilize the…
-
ThreatsDay: Android Spyware, PLC Attacks, AI Image Prompt Injection + 12 More Stories
Most of this week’s trouble came dressed as something useful.A package stole data. A fake extension opened remote access. A safety app became spyware. An image gave hidden orders to an AI agent. Other threats hid in open systems, weak code, and normal network traffic. The threats change every week. Subscribe, and we’ll alert you…
-
Mobile Synthetic Identity Scams Can Outscore Real Borrowers
Point Predictive’s Matt Vega on Detecting Identity Fraud and $30 Liveness Kits. Injection kits selling for about $30 enable fraud rings hot-wire a mobile device’s camera feed or inject a deepfake stream into the verification process, bypassing checks many institutions still rely on, said Matt Vega, chief fraud strategist at Point Predictive. First seen on…
-
Künstliche Intelligenz hat Hugging Face nicht an Cleverness übertroffen, sie hat das Mittelmaß optimiert
Die in der Diskussion stehende Attacke einer OpenAI-KI auf Hugging Face hat eine Zero-Day-Lücke in einem Package-Registry-Proxy ausgenutzt, um die Isolation zu durchbrechen. Die KI eskalierte in der Folge Privilegien, startete Seitwärtsbewegungen und erreichte so einen Knotenpunkt mit Internetzugang. Sie nutzte einen bösartigen Datensatz, um über unsichere Loader und Template-Injection die Ausführung von Remote-Code auszulösen.…
-
Microsoft Adds Prompt Injection Protection to Defender for Office 365
Microsoft has introduced prompt injection protection in Defender for Office 365, representing a significant advancement in securing enterprise email environments against emerging AI-targeted threats. As organizations increasingly adopt AI assistants like Microsoft 365 Copilot to summarize, triage, and respond to emails, attackers are shifting their tactics from traditional phishing methods to manipulating AI systems directly.…