We decided to create the AI Principles Map to help understand the trends, common threads, and differences among numerous sets of principles published.
You can;
- zoom in and out with mouse roll,
- click on pins to read the summaries,
- sort by date using the timeline on the right side,
- sort by type of organization using the checkbox at the bottom,
- search for specific documents using the search box on top,
- compare documents using the compare function,
- open this box by clicking on the button on the upper left corner.
The Map works best on Chrome desktop.
The Box is meant to help you visualize the ethical strengths and weaknesses of a technology. Once the weaknesses are identified, solutions can be designed!
Start your evaluation of a technology by checking how strongly it protects and promotes ethical values.
- Click on the core principles (autonomy, harm-benefit, justice) to unfold the related instrumental principles.
- Use the slider to rate the technology’s adherence to these principles.
- If an instrumental principle is not applicable, disable the slider using the N/A button.
For detailed information with a use-case, click How to use?
Documents:
0
1
2-4
5-8
9-16
17+
Compare AI principles
AI Principles by
Autonomy
- Power to decide (whether to decide)
- Human control
- Human oversight
- Transparency (to understand)
- Openness (to understand)
- Explainability
- Explicability
- Liberty
- Freedom
- Fundamental rights
- Personal privacy
- Privacy protection
- Fundamental rights
- Human values
No Harm
- Control risks
- Safety
- Security
- Capability caution
- Data protection
- Privacy (to avoid harm)
- Explicability
- Transparency (to avoid harm)
- Reproducibility
- Accuracy
- Reliability
- Responsible deployment
- Prevent arms race
Benefit
- Promoting well-being
- Benefit society
- Generating net benefits
- Sustaining the planet
- Impact
- Efficacy
- Explicability
- Scientific excellence
- User-centered design (for user benefit)
- People-first approach
Justice
- Fairness
- Fundamental rights
- Equality
- Non-discrimination
- Avoiding bias
- Inclusivity
- Diversity
- Data neutrality
- Representative data
- Shared benefit / prosperity
- Social & economic impacts
- Avoid disparity
- Mitigating social dislocation
- Preserving solidarity
- Accessibility
- Explicability
- Transparency (for accountability)
- Openness (for accountability)
- Accountability
- Auditability
- Liability
- Inclusive
- Judicial transparency
- Open governance
- Regulatory & legal compliance
Read Our Guide
None | Low | Medium | High | How to use? | ||
---|---|---|---|---|---|---|
Autonomy | ||||||
Harm-Benefit | ||||||
Justice |
Watch Our Video
Meet the Project Team
Cite Our Work
AI Ethics Lab,”Tool: The Box”, Toolbox: Dynamics of AI Principles, June 2020, http://aiethicslab.com/the-box/
Canca,“Operationalizing AI Ethics Principles”,Communications of the ACM, Volume 63, Number 12, 2020
Connected Works by Others
- Artificial Intelligence: The Global Landscape of Ethics Guidelines
- Principled Artificial Intelligence: Mapping Consensus in Ethical and Rights-based Approaches to Principles for AI
- Linking Artificial Intelligence Principles (table & paper)