FACTAA

Tracking since Apr 23, 2026.

Changelog

• May 29, 2026
  description

  Real-time verify ChatGPT-generated research citations and detect hallucinated references before they enter your paper.
  
  Verify ChatGPT citations before they become a problem
  AI tools like ChatGPT are now widely used for literature exploration and drafting research content. However, they often generate citations that look convincing but do not actually exist.
  These hallucinated references can easily slip into drafts — and may lead to serious issues during peer review, including credibility loss or manuscript rejection.
  This Chrome extension helps you detect and verify AI-generated citations instantly, directly inside ChatGPT.
  
  What this extension does
  Instead of manually checking every reference, the extension analyzes citations generated by ChatGPT and evaluates their validity in real time.
  Each reference is marked using a simple and intuitive traffic-light system:
  🟢 Verified — matches real publications and consistent metadata
  🟡 Suspicious — incomplete, inconsistent, or partially matched information
  🔴 Likely hallucinated — cannot be validated or may not exist
  This allows you to immediately identify problematic citations before using them in your work.
  
  How it works
  1.	Ask ChatGPT for research references
  2.	Click "Verify citations" in the extension
  3.	Instantly see verification results for each reference
  The extension integrates seamlessly into your workflow — no copy-pasting or external tools required.
  
  Key features
  Real-time citation verification: Analyze references directly within ChatGPT without leaving the page.
  Traffic-light evaluation system: Understand citation reliability at a glance using clear visual signals.
  Inline highlighting: Each reference is underlined and color-coded inside the ChatGPT response.
  Side panel overview: View all references in one place, along with their verification status.
  Direct source access: Click any reference to open the original source instantly.
  One-click citation copy: Easily copy verified references for use in your manuscript.
  
  Why this matters
  As AI becomes part of the research workflow, citation accuracy becomes harder to maintain.
  Hallucinated references are not just minor errors — they can:
  	undermine the credibility of your work
  	delay the review process
  	lead to rejection in serious cases
  This tool helps you catch these issues early, before they become a problem.
  
  Who should use this
  This extension is designed for researchers who rely on AI tools in their workflow:
  	Biomedical researchers
  	Professors and principal investigators
  	Postdoctoral researchers
  	Graduate students
  	Research scientists and analysts
  It is especially useful in fields where citation accuracy is critical, such as biomedical and life sciences.
  
  What makes this different
  Unlike generic reference tools, this extension focuses specifically on:
  	AI-generated citation verification
  	hallucination detection
  	seamless integration inside ChatGPT
  No manual lookup. No switching between tools. No guesswork.
  
  Important limitations
  This tool assists with citation verification but does not replace expert judgment.
  	Not all valid references may be detected
  	Some results may require manual confirmation
  	Users should always verify sources before submission
  
  Built for reliable AI-assisted research
  Use AI with confidence — without compromising academic integrity.
  
  Privacy & data handling     
  This extension follows a privacy-first design:
  	Only the text you explicitly select or submit is processed — the full conversation or page content is never read automatically.
  	Submitted text is processed transiently in server memory and is not permanently stored in our databases.
  	Only bibliographic metadata (DOI, PMID, title, authors) may be cached up to 30 days to improve re-verification speed.
  	External AI providers (OpenAI) are invoked with training-opt-out settings applied.
  	No personal data is sold, used for advertising, or shared with third parties beyond the service providers required to perform verification (Google OAuth, Crossref, NCBI, PubMed, arXiv, OpenAI).
  
  Supported platforms
  	chat.openai.com
  	chatgpt.com
  Support for additional platforms (Claude, Gemini, etc.) may be added in future versions with a separate review submission.
  
  Contact & support
  	Privacy policy: https://factaa.ai/privacy
  	Support email: [email protected]

  Real-time verify AI-generated research citations from ChatGPT and Claude, and detect hallucinated references before they enter your paper.
    
  Verify AI-generated citations before they become a problem
  AI tools like ChatGPT and Claude are now widely used for literature exploration and drafting research content. However, they often generate citations that look convincing
  but do not actually exist.
  These hallucinated references can easily slip into drafts — and may lead to serious issues during peer review, including credibility loss or manuscript rejection.
  This Chrome extension helps you detect and verify AI-generated citations instantly, directly inside ChatGPT and Claude.
  
  What this extension does
  Instead of manually checking every reference, the extension analyzes citations generated by ChatGPT and Claude and evaluates their validity in real time.
  Each reference is marked using a simple and intuitive traffic-light system:
  🟢 Verified — matches real publications and consistent metadata
  🟡 Suspicious — incomplete, inconsistent, or partially matched information
  🔴 Likely hallucinated — cannot be validated or may not exist
  This allows you to immediately identify problematic citations before using them in your work.
  
  How it works
  1.    Ask ChatGPT or Claude for research references
  2.    Click "Verify citations" in the extension
  3.    Instantly see verification results for each reference
  The extension integrates seamlessly into your workflow — no copy-pasting or external tools required.
  
  Key features
  Real-time citation verification: Analyze references directly within ChatGPT and Claude without leaving the page.
  Traffic-light evaluation system: Understand citation reliability at a glance using clear visual signals.
  Inline highlighting: Each reference is underlined and color-coded inside ChatGPT and Claude responses.
  Side panel overview: View all references in one place, along with their verification status.
  Direct source access: Click any reference to open the original source instantly.
  One-click citation copy: Easily copy verified references for use in your manuscript.
  
  Why this matters
  As AI becomes part of the research workflow, citation accuracy becomes harder to maintain.
  Hallucinated references are not just minor errors — they can:
      undermine the credibility of your work
      delay the review process
      lead to rejection in serious cases
  This tool helps you catch these issues early, before they become a problem.
  
  Who should use this
  This extension is designed for researchers who rely on AI tools in their workflow:
      Biomedical researchers
      Professors and principal investigators
      Postdoctoral researchers
      Graduate students
      Research scientists and analysts
  It is especially useful in fields where citation accuracy is critical, such as biomedical and life sciences.
   What makes this different
  Unlike generic reference tools, this extension focuses specifically on:
      AI-generated citation verification
      hallucination detection
      seamless integration inside ChatGPT and Claude
  No manual lookup. No switching between tools. No guesswork.
   Important limitations
  This tool assists with citation verification but does not replace expert judgment.
      Not all valid references may be detected
      Some results may require manual confirmation
      Users should always verify sources before submission
  
  Built for reliable AI-assisted research
  Use AI with confidence — without compromising academic integrity.
  
  Privacy & data handling
  This extension follows a privacy-first design:
      Only the text you explicitly select or submit is processed — the full conversation or page content is never read automatically.
      Submitted text is processed transiently in server memory and is not permanently stored in our databases.
      Only bibliographic metadata (DOI, PMID, title, authors) may be cached up to 30 days to improve re-verification speed.
      External AI providers (OpenAI) are invoked with training-opt-out settings applied.
      No personal data is sold, used for advertising, or shared with third parties beyond the service providers required to perform verification (Google OAuth, OpenAI,
  Crossref, NCBI PubMed E-utilities, arXiv, NANET National Assembly Library Search OpenAPI via data.go.kr, doi.org Content Negotiation).
  
  Supported platforms
      chat.openai.com
      chatgpt.com
      claude.ai
  Support for additional platforms (Gemini, etc.) may be added in future versions with a separate review submission.
  
  Contact & support
      Privacy policy: https://factaa.ai/privacy
      Support email: [email protected]

Permissions & access

Permissions

storagesidePanelidentitycontextMenusscripting

Host access

https://api-citation.ryxen.cloud/*, https://citation.ryxen.cloud/*, https://*.ryxen.cloud/*, https://accounts.google.com/*, https://www.googleapis.com/*

Screenshots