Large Reasoning Models Are Autonomous Jailbreak Agents

Jailbreaking -- bypassing built-in safety mechanisms in AI models -- has traditionally required complex technical procedures or specialized human expertise. In this study, we show that the persuasive capabilities of large reasoning models LRMs simplify and scale jailbreaking, converting it into a...

Proactive Disentangled Modeling of Trigger-Object Pairings for Backdoor Defense

Deep neural networks DNNs and generative AI GenAI are increasingly vulnerable to backdoor attacks, where adversaries embed triggers into inputs to cause models to misclassify or misinterpret target labels. Beyond traditional single-trigger scenarios, attackers may inject multiple triggers across...

PT-2025-31835

Name of the Vulnerable Software and Affected Versions Claude Code versions prior to 1.0.20 Claude Code versions prior to 1.0.24 Description Claude Code is an agentic coding tool. An error in command parsing allows bypassing the confirmation prompt, leading to the execution of untrusted commands...

CVE-2025-54132

Cursor is a code editor built for programming with AI. In versions below 1.3, Mermaid which is used to render diagrams allows embedding images which then get rendered by Cursor in the chat box. An attacker can use this to exfiltrate sensitive information to a third-party attacker controlled serve...

CVE-2025-54132 Cursor's Mermaid Diagram Tool is Vulnerable to an Arbitrary Image Fetch

Cursor is a code editor built for programming with AI. In versions below 1.3, Mermaid which is used to render diagrams allows embedding images which then get rendered by Cursor in the chat box. An attacker can use this to exfiltrate sensitive information to a third-party attacker controlled serve...

CVE-2025-54132

CVE-2025-54132 affects Cursor’s Mermaid-based diagram tool prior to version 1.3. An attacker can trigger prompt injections to cause Mermaid-rendered images to fetch data to a remote attacker-controlled server, enabling exfiltration of sensitive information. The issue is fixed in version 1.3; upgr...

CVE-2025-54132 Cursor's Mermaid Diagram Tool is Vulnerable to an Arbitrary Image Fetch

Cursor is a code editor built for programming with AI. In versions below 1.3, Mermaid which is used to render diagrams allows embedding images which then get rendered by Cursor in the chat box. An attacker can use this to exfiltrate sensitive information to a third-party attacker controlled serve...

CVE-2025-54132 Cursor's Mermaid Diagram Tool is Vulnerable to an Arbitrary Image Fetch

Cursor is a code editor built for programming with AI. In versions below 1.3, Mermaid which is used to render diagrams allows embedding images which then get rendered by Cursor in the chat box. An attacker can use this to exfiltrate sensitive information to a third-party attacker controlled serve...

CVE-2025-54131 Cursor bypasses its allow list to execute arbitrary commands

Cursor is a code editor built for programming with AI. In versions below 1.3, an attacker can bypass the allow list in auto-run mode with a backtick or $cmd. If a user has swapped Cursor from its default settings requiring approval for every terminal call to an allowlist, an attacker can execute...

CVE-2025-54131 Cursor bypasses its allow list to execute arbitrary commands

Cursor is a code editor built for programming with AI. In versions below 1.3, an attacker can bypass the allow list in auto-run mode with a backtick or $cmd. If a user has swapped Cursor from its default settings requiring approval for every terminal call to an allowlist, an attacker can execute...

Cursor AI Code Editor Fixed Flaw Allowing Attackers to Run Commands via Prompt Injection

Cybersecurity researchers have disclosed a now-patched, high-severity security flaw in Cursor, a popular artificial intelligence AI code editor, that could result in remote code execution RCE. The vulnerability, tracked as CVE-2025-54135 CVSS score: 8.6, has been addressed in version 1.3 released...

Cursor AI Code Editor Fixed Flaw Allowing Attackers to Run Commands via Prompt Injection

Cybersecurity researchers have disclosed a now-patched, high-severity security flaw in Cursor, a popular artificial intelligence AI code editor, that could result in remote code execution RCE. The vulnerability, tracked as CVE-2025-54135 CVSS score: 8.6, has been addressed in version 1.3 released...

CVE-2025-54131: Improper Neutralization of Special Elements used in a Command

Cursor is a code editor built for programming with AI. In versions below 1.3, an attacker can bypass the allow list in auto-run mode with a backtick or $cmd. If a user has swapped Cursor from its default settings requiring approval for every terminal call to an allowlist, an attacker can execute...

CVE-2025-54132: Server-Side Request Forgery (SSRF)

Cursor is a code editor built for programming with AI. In versions below 1.3, Mermaid which is used to render diagrams allows embedding images which then get rendered by Cursor in the chat box. An attacker can use this to exfiltrate sensitive information to a third-party attacker controlled serve...

PT-2025-31700

Name of the Vulnerable Software and Affected Versions Cursor versions prior to 1.3 Description Cursor is a code editor built for programming with AI. In versions below 1.3, Mermaid, a tool used for rendering diagrams, allows embedding images. An attacker can leverage this functionality to...

PT-2025-31699 · Cursor · Cursor

Name of the Vulnerable Software and Affected Versions: Cursor versions prior to 1.3 Description: Cursor, a code editor built for programming with AI, allows an attacker to bypass the allow list in auto-run mode using a backtick or $cmd. This bypass enables arbitrary command execution outside of t...

Browser Extensions Can Exploit ChatGPT, Gemini in ‘Man in the Prompt’ Attack

Man in the Prompt attack shows how browser extensions can exploit ChatGPT, Gemini and other AI tools to steal data or inject hidden prompts...

CVE-2025-43252

This issue was addressed by adding an additional prompt for user consent. This issue is fixed in macOS Sequoia 15.6. A website may be able to access sensitive user data when resolving symlinks...

CVE-2025-43184

This issue was addressed by adding an additional prompt for user consent. This issue is fixed in macOS Sonoma 14.7.7, macOS Ventura 13.7.7, macOS Sequoia 15.4. A shortcut may be able to bypass sensitive Shortcuts app settings...

SAEL: Leveraging Large Language Models with Adaptive Mixture-Of-Experts for Smart Contract Vulnerability Detection

With the increasing security issues in blockchain, smart contract vulnerability detection has become a research focus. Existing vulnerability detection methods have their limitations: 1 Static analysis methods struggle with complex scenarios. 2 Methods based on specialized pre-trained models...