Hugging Face's logo

Datasets:
hewyler
/
AILossTaxonomy

Tasks:
Tabular Classification
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
< 1K
Tags:
GRC
AI-governance
risk-management
quantitative-analysis
finops
Libraries:
Datasets
pandas
License:
Dataset card Data Studio Files
xet
Community
Dataset Viewer
Auto-converted to Parquet Duplicate
Domain
stringclasses
4 values
Loss Name
stringlengths
14
24
Explanation
stringlengths
147
173
Source
stringclasses
1 value
Compliance
Legal Compensations
Settlement payments to affected parties for harm caused by AI malfunctions or decisions. Attorney fees and court costs for defending lawsuits from individuals or groups.
Prof. Hernan Huwyler
Compliance
Contractual Credits
Service credits issued to customers when AI performance falls below guaranteed levels (SLAs). Refunds and discounts applied for missed availability or accuracy commitments.
Prof. Hernan Huwyler
Compliance
Regulatory Fines
Penalties for violating AI regulations like EU AI Act, GDPR, or NYC 144. Sanctions for data breaches, discriminatory outcomes, or copyright infringements.
Prof. Hernan Huwyler
Compliance
Legal Response
External legal counsel fees for investigating and responding to AI-related claims. Internal legal team costs for compliance reviews and regulatory correspondence.
Prof. Hernan Huwyler
Compliance
Control Remediation
Costs to fix governance gaps identified in failed AI audits. Documentation, implementation, and certification expenses for new compliance controls and frameworks.
Prof. Hernan Huwyler
IT/Technical
Data Regeneration
Costs to rebuild training datasets when data becomes corrupted, poisoned, or drifted. Expenses for new data collection, labeling, cleaning, and validation.
Prof. Hernan Huwyler
IT/Technical
Algorithm Remediation
Engineering costs to retrain models that produce biased or inaccurate predictions. Compute resources and testing expenses for fixing drifted or poorly performing algorithms.
Prof. Hernan Huwyler
IT/Technical
Infrastructure Overruns
Unexpected cloud computing (GPU/TPU) and storage costs from inefficient AI resource usage. Emergency scaling expenses when systems face performance bottlenecks.
Prof. Hernan Huwyler
Operational
Decision Errors
Financial losses from incorrect AI-driven business decisions made at scale (e.g., bad loans, wrong inventory purchases). Costs of resource misallocation.
Prof. Hernan Huwyler
Operational
Operational Inefficiency
Manual intervention costs when humans must correct or override AI outputs (Human-in-the-loop costs). Lost productivity from rework and staff time diverted.
Prof. Hernan Huwyler
Operational
Development Waste
Write-off of failed AI projects that never reach production deployment. Sunk costs in licenses, development efforts, and procurement that yield no value.
Prof. Hernan Huwyler
Operational
Business Disruption
Revenue loss during downtime when AI-dependent processes stop functioning. Emergency replacement costs and lost transactions from service interruptions.
Prof. Hernan Huwyler
Operational
Provider Switching
Contract termination fees and cancellation penalties with current AI vendors/LLMs. Migration costs, integration expenses, and negotiation time for new provider onboarding.
Prof. Hernan Huwyler
Revenue
Customer Churn
Lost revenue from customers leaving after negative AI experiences (hallucinations, poor chatbots). Acquisition costs for replacing churned clients.
Prof. Hernan Huwyler
Revenue
Reputation Damage
Brand value decline and crisis management costs following publicized AI incidents. Lost business opportunities and reduced market position from negative media coverage.
Prof. Hernan Huwyler
YAML Metadata Warning: The task_categories "risk-assessment" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other

AI Loss Taxonomy & Financial Impact Framework

Dataset Summary

This dataset provides a structured taxonomy of AI Loss Events, designed to help organizations quantify the financial impact of Artificial Intelligence failures. Unlike threat lists which focus on probability, this taxonomy focuses on magnitude (impact), enabling Quantitative Risk Analysis (e.g., FAIR methodology) for AI systems.

It categorizes losses across Compliance, IT/Technical, Operational, and Revenue domains, providing a clear vocabulary for CFOs, CROs, and CISOs to discuss AI risk.

Author & Attribution

This taxonomy was developed and curated by: Prof. Hernan Huwyler, MBC, CPA

  • Academic Director
  • AI GRC Director

Please attribute Prof. Hernan Huwyler when utilizing this taxonomy for academic or commercial risk frameworks.

Dataset Structure

The dataset contains the following fields:

  • Domain: The business area where the financial loss materializes (e.g., Compliance, Revenue).
  • Loss Name: Standardized terminology for the specific type of financial loss.
  • Explanation: Detailed definition of the loss, including direct and indirect costs.
  • Source: Attribution field.

Use Cases

1. Quantitative Risk Analysis (FAIR)

Use this dataset to define the Loss Magnitude side of your risk equation.

  • Scenario: An AI model exhibits bias (Threat).
  • Loss Calculation: Use "Regulatory Fines" + "Reputation Damage" + "Algorithm Remediation" from this dataset to calculate the Total Loss Exposure (TLE).

2. AI ROI & Cost-Benefit Analysis

When calculating the Return on Investment for AI, use the "Operational" and "IT/Technical" loss categories to estimate potential "Development Waste" and "Infrastructure Overruns" to calculate a risk-adjusted ROI.

3. Insurance & Liability

This taxonomy assists insurance actuaries and corporate risk managers in defining coverage limits for AI liability policies by identifying specific cost drivers like "Legal Response" vs "Data Regeneration."

Example Data

Domain Loss Name Explanation
Compliance Regulatory Fines Penalties for violating AI regulations like EU AI Act or privacy laws.
IT/Technical Algorithm Remediation Engineering costs to retrain models that produce biased or inaccurate predictions.
Revenue Customer Churn Lost revenue from customers leaving after negative AI experiences.

Citation

If you use this dataset in research or tooling, please cite:

Huwyler, H. (2025). AI Loss Taxonomy. Hugging Face Datasets.

Downloads last month
15
Size of downloaded dataset files:
4.78 kB
Size of the auto-converted Parquet files:
4.78 kB
Number of rows:
15

Space using hewyler/AILossTaxonomy 1

Collection including hewyler/AILossTaxonomy