EthicsNet Creed.Space
EthicsNet Creed.Space is a nonprofit initiative led by Nell Watson that works on two interconnected fronts: EthicsNet crowdsources datasets of prosocial behaviors from people worldwide to teach AI systems human values and cultural norms, modeled after ImageNet's approach to computer vision. Creed.Space extends this into a practical constitutional AI safety platform where users select from 113+ ethical 'creeds' (machine-readable value specifications) to shape AI behavior with adjustable strictness levels. The Personalized Constitutionally-Aligned Agentic Superego framework integrates with major AI models via Model Context Protocol, achieving a 96.4% reduction in harmful AI outputs in published benchmarks.
Funding Details
- Annual Budget
- -
- Monthly Burn Rate
- -
- Current Runway
- -
- Funding Goal
- -
- Funding Raised to Date
- -
- Fiscal Sponsor
- Players Philanthropy Fund
Theory of Change
EthicsNet Creed.Space believes that AI systems need large, well-documented datasets of prosocial behaviors to learn ethical conduct, much as ImageNet provided the data foundation for computer vision breakthroughs. By crowdsourcing annotations of prosocial behavior from people across cultures, they create training data that reflects diverse human values. The Creed.Space platform then operationalizes this into practical safety infrastructure: machine-readable ethical specifications ('creeds') that can be applied as constitutional guardrails on any AI system. Rather than relying on any single AI company's alignment choices, this approach enables personalized, verifiable, culturally-aware alignment that persists across deployments and operators. The causal chain runs from community-sourced ethical data to standardized ethical specifications to real-time AI behavior oversight, reducing harmful outputs and enabling AI systems to respect diverse human values.
Grants Received
from Survival and Flourishing Fund
Projects
No linked projects.
People
No linked people.
Discussion
Sign in to join the discussion.
No comments yet. Be the first to share your thoughts.
Details
- Last Updated
- Apr 2, 2026, 9:48 PM UTC
- Created
- Mar 18, 2026, 11:18 PM UTC