LangSmith Client SDKs 代码问题漏洞

LangSmith Client SDKs are a developer toolkit open-sourced by LangChain. Versions of LangSmith Client SDKs prior to 0.8.0 and JS/TS versions prior to 0.6.0 have code vulnerabilities. This vulnerability stems from the lack of differentiation between public prompts and internal organization-specifi...

Defending Large Language Models against Jailbreak Exploits with Responsible AI Considerations

Large Language Models LLMs remain susceptible to jailbreak exploits that bypass safety filters and induce harmful or unethical behavior. This work presents a systematic taxonomy of existing jailbreak defenses across prompt-level, model-level, and training-time interventions, followed by three...

GPT Academic 命令注入漏洞

GPT Academic is an interface that provides pragmatic interactions for LLM grand language models such as GPT/GLM. GPT Academic suffers from a command injection vulnerability that stems from improper handling of user-supplied prompts in the manim plugin, which can be exploited by an attacker to cau...

Important: thunderbird

Issue Overview: The Mozilla Foundation Security Advisory describes this flaw as: It was possible for certain browser prompts and dialogs to be activated or dismissed unintentionally by the user due to an insufficient activation-delay. CVE-2023-5721 The Mozilla Foundation Security Advisory describ...

DEBIAN-CVE-2023-4901

Inappropriate implementation in Prompts in Google Chrome prior to 117.0.5938.62 allowed a remote attacker to potentially spoof security UI via a crafted HTML page. Chromium security severity: Medium...

SUSE CVE-2023-2459

Inappropriate implementation in Prompts in Google Chrome prior to 113.0.5672.63 allowed a remote attacker to bypass permission restrictions via a crafted HTML page. Chromium security severity: Medium...

CVE-2020-10595

A flaw was found during prompting initiated by the Kerberos library, where an attacker who enters a response exactly as long as the length of the buffer provided by the underlying Kerberos library, causes pam-krb5 to write a single null byte past the end of that buffer. This flaw results in heap...

CVE-2018-6160

JavaScript alert handling in Prompts in Google Chrome prior to 68.0.3440.75 allowed a remote attacker to spoof the contents of the Omnibox URL bar via a crafted HTML page...