AI Risk Repository

Submit an AI Risk, Incident, or Concern

Confidential • Responsible • Research-driven

As part of the AI4Food research initiative, we have developed an AI Risk Repository for Agriculture and Food Systems.
The purpose of this repository is to systematically identify, document, and analyze risks associated with the use of AI across the agri-food sector.

Our intention is not only to list risks, but to:

  • Understand how AI risks emerge in real agricultural contexts

  • Support the development of practical mitigation strategies

  • Help protect farmers, farm data, and food systems

  • Enable organizations to operationalize responsible AI

  • Inform policy, governance frameworks, and best practices

This repository is designed as a living resource, evolving with new technologies, use cases, and evidence from research and practice. It is a collaborative effort. By contributing, you help ensure that AI in agriculture and food systems is safe, fair, transparent, and beneficial for all.

Why Contribute?

AI systems can create unintended risks across safety, fairness, privacy, security, and misuse. By sharing observed risks or incidents, you contribute to a growing evidence base that supports:

01

Better AI governance and oversight

02

Improved risk mitigation practices

03

Informed policy and standards development

The AI Risk Repository helps different stakeholders:

For farmers

  • Understand how AI tools may impact farm operations, data, and decision-making

  • Identify risks early and ask informed questions

For ag-tech providers (ATPs)

  • Anticipate technical, ethical, and operational risks

  • Improve system design, documentation, and governance practices

For policymakers and researchers

  • Identify recurring and systemic AI risks

  • Support evidence-based policy development and regulation

For the broader ecosystem

  • Promote transparency, accountability, and shared learning

  • Avoid repeating known failures or harmful practices

Share an AI Risk from Research or Practice

We invite farmers, researchers, technology providers, policymakers, and practitioners to contribute to this repository by sharing identified, observed, or potential AI risks in agriculture and food systems.

You can submit:

  • Risks identified through academic literature

  • Risks observed during technology deployment

  • Risks emerging from policy, regulatory, or governance gaps

  • Risks affecting farmers, workers, consumers, or ecosystems

Submit a Risk Entry

 Please complete the form below to submit an AI risk entry

AI Risk Taxonomy

A Structured Way to Understand AI Risks

To ensure consistency and clarity, we have developed a structured AI Risk Taxonomy tailored to agriculture and food systems.

This taxonomy helps:

  • Categorize risks across the AI lifecycle

  • Distinguish between technical, data, ethical, environmental, legal, and governance risks

  • Capture sector-specific impacts unique to farming and food production

References and Foundations

The taxonomy has been developed with reference to established global frameworks, including:

  • The MIT AI Risk Repository

  • The MITRE AI Risk Framework

Below is the AI Risk Taxonomy used in this repository. Each submitted risk is mapped to this structure to ensure consistency and comparability.

Confidentiality and Data Use

  • Submissions are treated confidentially

  • Data is used solely for AI risk analysis, research, and governance activities

  • Identifiable information will not be published without explicit consent

  • Data may be anonymized and aggregated for insights and reporting

 

How This Repository Is Used

  • All submissions are reviewed and categorized using the taxonomy

  • Risks are documented to support analysis, comparison, and learning

  • Insights from the repository feed into:

    • AI governance toolkits

    • Risk mitigation guidance

    • Policy and regulatory discussions

    • Capacity-building for farmers and ATPs