EthicsBench

AI Ethics and Regulation, Structured.EthicsBench is a neutral and open-access resource that curates leading ethical benchmarks and AI governance frameworks from around the world.Designed for developers, researchers, educators, and policymakers who seek clarity in a rapidly evolving field.

Key Ethical Benchmarks- ETHICS Dataset
Real-world moral judgments from a wide variety of contexts.- SCRUPLES
Crowdsourced moral dilemmas and choices in everyday situations.- Moral Machine
Ethical scenarios for autonomous vehicles — a global experiment.- IEEE EAD
Ethically Aligned Design principles for responsible AI systems.→ More benchmarks to be added soon.

AI Regulations Overview- EU AI Act (2024)
The first comprehensive legal framework for AI in the European Union.- U.S. AI Bill of Rights
A rights-based framework for algorithmic accountability in the United States.- OECD AI Principles
International standards promoting trustworthy AI.- Japan AI Guidelines
Emphasizing human-centered AI development and use.→ Links to official documents provided below.

EU AI Act (2024)
Regulatory framework for trustworthy AI in the EU.U.S. AI Bill of Rights
Rights-based framework for protecting the public in the age of AI.OECD AI Principles
Global principles to foster responsible AI innovation.SCRUPLES Dataset (AllenAI)
A corpus of ethical dilemmas and public responses.ETHICS Dataset (Hendrycks)
Real-world ethical decision-making scenarios for evaluating AI behavior.

AI Strategy 2022 – Cabinet Office (English)
National vision for a human-centric AI society.AI Guidelines for Business – METI (English)
Practical guidance for ethical and secure AI use in industry.AI 利活用ガイドライン Ver1.0（日本語）
企業向けのAI利活用に関する日本語ガイドライン。Japan's AI Principles Overview (CSIS)
国際的な文脈における日本の倫理原則の位置づけ。

Resources- Official documents (PDFs, frameworks, legal texts)
- Comparative tables (in progress)
- Research papers and datasets
- Ethics evaluation benchmarks→ A shared foundation for those navigating AI’s ethical and regulatory terrain.

About EthicsBenchEthicsBench is not a think tank or opinion platform.
It offers a neutral map of the existing ethical benchmarks and regulations shaping the future of AI.We do not recommend.
We organize.
We believe structure is the first step to understanding.This project is independently maintained.