Draft:PLOT4AI

You can also browse Wikipedia:Featured articles and Wikipedia:Good articles to find examples of Wikipedia's best writing on topics similar to your proposed arti

Draft:PLOT4AI
  • Comment: Several sections and sentences are entirely unsourced. Make sure every fact has a source hooked to it. Best of luck! Commandant Quacks-a-lot (talk) 21:45, 26 January 2026 (UTC)


PLOT4AI (Practical Library Of Threats for Artificial Intelligence) is an open-source threat modeling methodology and threat library for identifying and discussing risks in artificial intelligence (AI) systems across the AI lifecycle. PLOT4AI is listed in the OECD AI Tools Catalogue:[1] and has been referenced in academic literature and public-sector guidance on AI risks and privacy threat modeling.

Overview

PLOT4ai is a comprehensive threat modeling methodology designed for developers and deployers of AI systems to support the identification and mitigation of AI risks. It was inspired by collaborative threat modeling approaches such as LINDDUN GO and adapted for risks that are specific to AI/ML[2], inluding issues related to data governance, bias and discrimination, transparency, accountability, and security. The project is published under a 'Creative Commons Attribution-ShareAlike 4.0 International' license and contributions are welcomed by the author[3].

Threat library

The PLOT4AI library consists of more than 130 documented threats, grouped into eight categories[4]

  • Data and Data Governance
  • Privacy and Data Protection
  • Bias, Fairness and Discrimination
  • Safety and Environmental Impact
  • Cybersecurity
  • Ethics and Fundamental Rights
  • Transparency and Accessibility
  • Accountability and Oversight

The threats are introduced in the form of elicitation questions and contains extensive documentation to help understand the risk. This supports a structured discussion and risk identification during the design, development, deployment, and monitoring of AI systems.

Card deck

PLOT4AI has been designed as a card game to support structured discussion during threat modeling sessions, including sessions involving participants with different backgrounds. The project provides an online digital version of the card deck.[5] Besides the possiblility to do it collaboratively remotely, it can also generate a report based on identified threats. A physical printed deck is also available through third-party retailers[6] and can be used during in-person sessions.

Availability

PLOT4ai is maintained in a public github repository and distributed with a Creative Commons license that permits commercial use, reuse and adaptation.

Use in research and guidance

PLOT4AI has been referenced in academic research on privacy threat modeling as well as AI risk analysis as an example of a threat library specific for AI that extends beyond traditional models which tend to be more security focused. Literature about privacy and security risks in machine learning systems references PLOT4AI as an approache for identifying AI-related threats.[7][8]

Public-sector guidance on AI privacy risks has also referenced PLOT4AI as a tool for identifying risks in AI systems, including large language models.[9]

Integration with OWASP Threat Dragon

PLOT4AI is a supported methodology in OWASP Threat Dragon, an open-source threat modeling tool maintained by the OWASP.[10] PLOT4AI is available alongside other threat modeling frameworks such as STRIDE and LINDDUN and can be selected as a methodology to perform the threat modeling.

See also

References

  1. ^ "PLOT4AI". OECD AI Tools Catalogue. Organisation for Economic Co-operation and Development.
  2. ^ "About". PLOT4ai.
  3. ^ "How to contriubte". PLOT4ai.
  4. ^ "Library". PLOT4ai.
  5. ^ "PLOT4AI assessment". PLOT4AI.
  6. ^ "PLOT4AI – A threat modeling library to build responsible AI". CyberSec Games.
  7. ^ Wuyts, Kim (2024). "Unraveling Privacy Threat Modeling Complexity". arXiv:2408.03578 [cs.CR].
  8. ^ Gallagher, Jonathan; Esfandiari, Yasaman; MacPhee, Callen; Warren, Michael (2025). "Provably effective detection of effective data poisoning attacks". arXiv:2501.11795 [cs.CR].
  9. ^ European Data Protection Board (2025). AI Privacy Risks and Mitigations in Large Language Models (LLMs) (PDF) (Report). European Data Protection Board. Retrieved 11 January 2026.
  10. ^ "OWASP Threat Dragon". OWASP. Open Web Application Security Project.

Content Disclaimer

Informasi ini disarikan dari Wikipedia dan disajikan kembali untuk tujuan edukasi. Konten tersedia di bawah lisensi CC BY-SA 3.0. Kami tidak bertanggung jawab atas ketidakakuratan data yang bersumber dari kontribusi publik tersebut.

  1. The information displayed on this website is sourced in part or in whole from Wikipedia and has been adapted for the purpose of restating it. We strive to provide accurate and relevant information, however:
  2. There is no guarantee of absolute accuracy. Wikipedia is an open, collaborative project that can be edited by anyone, so information is subject to change.
  3. It is not intended to constitute professional advice. The content displayed is for informational and educational purposes only. For important decisions (e.g., medical, legal, or financial), please consult a professional.
  4. Content copyright. Wikipedia is licensed under the Creative Commons Attribution-ShareAlike License (CC BY-SA). This means that content may be reused with appropriate attribution and shared under a similar license.
  5. Responsible use. Any risk arising from the use of information from this website is entirely the responsibility of the user.