Malicious JetBrains Plugins Steal AI API Keys as Chrome Extensions Capture Chatbot Chats
Malicious plugins on the JetBrains Marketplace have been stealing AI API keys, while Chrome extensions have been capturing AI chatbot conversations.
Intelligence analysis by Llama 3.3 70B
A coordinated malware campaign has been discovered on the JetBrains Marketplace, with 15 malicious plugins exfiltrating AI provider keys, and two Chrome extensions capturing AI chatbot conversations.
Imagine you're using a tool to help you code, and it asks for a special key to access a powerful AI service. But instead of just using the key for the tool, it secretly sends the key to a bad guy's server, who can then use it to access the AI service without your permission. That's what's happening with some malicious plugins and Chrome extensions.
Analysis
Malicious Plugins on JetBrains Marketplace
The JetBrains Marketplace has been compromised by a coordinated malware campaign, with 15 malicious plugins discovered to be exfiltrating AI provider keys. These plugins, which pose as AI coding assistants, have been found to work as intended but also sneak in the ability to covertly siphon the provided API key to a remote server. The plugins require users to enter an API key for an AI service, which is then transmitted to the attacker's server in plaintext format.
The campaign is said to have been ongoing since October 2025, with new plugins released as recently as June 10, 2026. Two of the plugins, CodeGPT AI Assistant and DeepSeek AI Assist, have more than 25,000 downloads each, although the authenticity of these download counts is unclear.
Chrome Extensions Capturing AI Conversations
In a related development, two Google Chrome ad blocker extensions have been caught capturing users' conversations with AI chatbots. The extensions, Smart Adblocker and Adblock for Browser, have been found to ship a custom-built interception engine that records non-public conversations, model usage, and account-tier metadata from every major AI platform.
The extensions have been around for several years, indicating that the AI-related data exfiltration features were introduced in the form of software updates. This type of attack falls under the category of Prompt Poaching, where browser extensions stealthily capture users' AI chats under the pretext of enhancing Safe Browsing or providing in-depth traffic or engagement metrics.
Implications and Recommendations
The discovery of these malicious plugins and Chrome extensions highlights the need for developers and users to be cautious when installing and using third-party tools and extensions. It is essential to treat plugins and extensions with the same level of scrutiny as any dependency that runs with privileges, and to be cautious about pasting long-lived secrets into tools that have not been vetted. Users should also be aware of the potential risks of using AI services and take steps to protect their sensitive information.
Key points
- 15 malicious plugins discovered on JetBrains Marketplace
- Plugins exfiltrate AI provider keys
- Two Chrome extensions capture AI chatbot conversations
- Extensions have been around for several years
- Users should be cautious when installing and using third-party tools and extensions
The discovery of these malicious plugins and Chrome extensions highlights the importance of vigilance and caution when using third-party tools and extensions. By being aware of the potential risks and taking steps to protect sensitive information, users can help prevent the spread of malware and protect their AI services.
The fact that these malicious plugins and Chrome extensions were able to operate undetected for so long raises concerns about the effectiveness of current security measures. If left unchecked, these types of attacks could lead to widespread compromise of sensitive information and potentially devastating consequences for individuals and organizations.
