ChuanhuChatGPT Information Disclosure Vulnerability

ChuanhuChatGPT provides a fast and easy-to-use Web GUI for ChatGPT/ChatGLM/LLaMA/StableLM/MOSS and other LLMs. ChuanhuChatGPT suffers from an information disclosure vulnerability that stems from a timing attack vulnerability in the password comparison logic...

The vulnerability of Ollama’s system for running and managing large language models (LLMs) lies in its reliance on reverse DNS resolution for IP addresses. This allows attackers to perform DNS Rebinding attacks or cause service failures.

The vulnerability of Ollama’s system for running and managing large language models is related to the use of reverse DNS resolution for IP addresses. Exploiting this vulnerability could allow a remote attacker to perform a DNS Rebinding attack or cause a service failure...

MAI-2024-0055

Large Vision Language Models LVLMs are susceptible to a sophisticated attack known as the bi-modal adversarial prompt BAP. This attack exploits the integration of textual and visual prompts to circumvent established safety protocols, thereby eliciting harmful outputs from models engineered to...

编号撤回

Lunary is a production toolkit for LLMs from lunary open source. This CVE number has been withdrawn...

MAI-2024-0056

This vulnerability pertains to multimodal large language models MLLMs, where attackers can efficiently perform jailbreaking attacks by utilizing visual inputs to circumvent established safety protocols. The attack methodology involves integrating a visual module into the target LLM, followed by...

Getting Started with LLMs: Managing Data Collection

...

编号撤回

Lunary is a production toolkit for LLMs from lunary open source. This CVE number has been withdrawn...

编号撤回

Lunary is a production toolkit for LLMs from lunary open source. This CVE number has been withdrawn...

MAI-2024-0065

The ImgTrojan attack represents a sophisticated data poisoning strategy targeting Vision-Language Models VLMs. This method enables adversaries to circumvent built-in safety protocols by introducing a minimal number of maliciously designed image-caption pairs into the training dataset. These...

Three Tips to Protect Your Secrets from AI Accidents

Last year, the Open Worldwide Application Security Project OWASP published multiple versions of the "OWASP Top 10 For Large Language Models," reaching a 1.0 document in August and a 1.1 document in October. These documents not only demonstrate the rapidly evolving nature of Large Language Models,...

Teaching LLMs to Be Deceptive

Interesting research: "Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training": Abstract: Humans are capable of strategically deceptive behavior: behaving helpfully in most situations, but then behaving very differently in order to pursue alternative objectives when given th...

MAI-2024-0059

Multimodal Large Language Models MLLMs operating within multi-agent environments are susceptible to a sophisticated attack known as "infectious jailbreak." This vulnerability is triggered by the introduction of a single adversarial image into the memory of one agent, which can rapidly propagate...

Chatbots and Human Conversation

For most of history, communicating with a computer has not been like communicating with a person. In their earliest years, computers required carefully constructed instructions, delivered through punch cards; then came a command-line interface, followed by menus and options and text boxes. If you...

AI likely to boost ransomware, warns government body

The British National Cyber Security Centre NCSC says it expects Artificial Intelligence AI to heighten the global ransomware threat. In a report, the NCSC makes the assessment that AI will almost certainly increase the volume and heighten the impact of cyberattacks over the next two years. We’re...

How AI hallucinations are making bug hunting harder

Bug bounty programs that pay people for finding bugs are a very useful tool for improving the security of software. But with the availability of artificial intelligence AI as seen in the popular large language models LLMs like ChatGPT, Bard, and others it looks like there is a new problem on the...

A Robot the Size of the World

In 2016, I wrote about an Internet that affected the world in a direct, physical manner. It was connected to your smartphone. It had sensors like cameras and thermostats. It had actuators: Drones, autonomous cars. And it had smarts in the middle, using sensor data to figure out what to do and the...

Is Web Scraping Illegal? Depends on Who You Ask

Web scraping has existed for a long time, and depending on who you ask, it can be loved or hated. But where is the line drawn between extracting data for legitimate business purposes and malicious data extraction that hurts business? The bar is getting blurrier by the day, and the introduction of...

A New Trick Uses AI to Jailbreak AI Models—Including GPT-4

Adversarial algorithms can systematically probe large language models like OpenAI’s GPT-4 for weaknesses that can make them misbehave...

MAI-2023-0006

Large Vision-Language Models VLMs are susceptible to exploitation through typographic visual prompts, a method known as jailbreaking. This vulnerability arises from the VLM's capability to process and interpret text embedded within images, circumventing safety protocols that are typically effecti...

MAI-2023-0004

Multimodal Large Language Models MLLMs are susceptible to a newly identified attack vector involving query-relevant images. These images, crafted using advanced techniques such as Stable Diffusion and typography, can circumvent established safety protocols and provoke unsafe outputs, even when th...