Is "AI Chat Exporter" on Chrome Web Store Safe to Install?

AI Security Review

Risky Plugins reviewed this extension with an AI-assisted security workflow on 2026-05-05. The review verdict is benign but powerful with 70% confidence.

Recommended action: no action.
Risk context: LOW risk, score 31/100.
Evidence context: threat category none; evidence quality moderate.

The AI Chat Exporter extension demonstrates a clean security profile with minimal findings that align with expected functionality for a content injection tool. The extension's sole network finding—a fetch call detected in platforms/claude/injector.js:6—represents standard content script behavior for an extension designed to interact with AI chat platforms. This file path structure indicates a modular architecture organized by platform (Claude, with similar files likely for ChatGPT, Gemini, and Copilot based on the description), which is consistent with the extension's stated purpose of exporting conversations from multiple AI services.

No malware signatures were detected across the codebase, and there are zero obfuscation findings. The absence of code-smell findings further indicates clean, maintainable code. Importantly, there are no suspicious IoCs or external domains beyond what would be expected for the extension's legitimate functionality. The findings summary shows 0 IoCs, 0 malware signatures, 0 obfuscation indicators, and 0 code-smell findings—the only finding is the single medium-severity network finding for a fetch call.

The developer email ([email protected]) and zero user count represent weak metadata signals, but these do not constitute security concerns. The extension's name and description are coherent and match the expected behavior of a chat export tool. The version number (1.8.2) suggests iterative development rather than a hastily deployed malicious extension.

A skeptic might argue that the zero user count and Gmail developer email suggest this could be a newly deployed extension with unverified intentions. However, the code analysis reveals no evidence of malicious behavior. The single network finding is a fetch call, which is necessary for content scripts to interact with web pages and extract conversation data. Without malware signatures, obfuscation, or suspicious domains, there is no basis for concern. The file platforms/claude/injector.js follows a logical naming convention for platform-specific content injection code.

The extension's capabilities (network access, content injection) could theoretically be misused for data exfiltration, but the current implementation shows no such patterns. The clean findings profile and legitimate functionality suggest this is a benign extension with normal capabilities. Recommended action is no_action, as there is no evidence of malicious behavior requiring intervention.

Key Reasons

• Single network finding is expected fetch call for content injection in platforms/claude/injector.js
• No malware signatures or obfuscation detected across codebase
• Coherent name and description match expected export functionality
• No suspicious IoCs or external domains beyond legitimate use
• File structure suggests legitimate modular content script architecture

False Positive Considerations

• Network finding is benign fetch call expected for content scripts