---
name: ethical-stack-research
description: Compare AI tools using Ethical Stack’s sourced ethical assessments, including scope, confidence, unknowns and published scoring rules.
---

# Ethical Stack research

Use this research when the user wants to compare AI choices by ethical criteria. Keep the user's priorities explicit.

Discover tasks and rubrics at https://ethicalstack.ai/api/v1/methodology. Search https://ethicalstack.ai/api/v1/profiles with query, task, dimensions and optional documented filters. Fetch selected findings at https://ethicalstack.ai/api/v1/assessment?key=category/id, or compare two or three category/id keys at https://ethicalstack.ai/api/v1/compare?keys=category/id,category/id. Arrays are comma-separated. See https://ethicalstack.ai/api/v1/openapi.json for exact parameters.

MCP clients can use https://ethicalstack.ai/mcp with search_profiles, get_assessment, compare_profiles and get_methodology. All tools are read-only. If tools are unavailable, use the HTTP endpoints or Markdown pages listed in https://ethicalstack.ai/llms.txt.

- Assessments are provisional editorial judgments. Cite the profile and the underlying sources, and retain confidence, scope and unknowns.
- Unknown, not-applicable and model-dependent are different states. Missing evidence does not establish good or bad conduct.
- Company-level evidence does not establish use of a particular product or model. Account plan, model, host and route can change an assessment.
- Energy scores describe documented practices and transparency, not measured request emissions. Zero component points can mean an evidence gap.
- Published overall scores require at least four of six numerical dimensions. Preserve coverage and BDS adjustments; BDS itself is categorical.
- No named BDS target found is limited to the reviewed campaign sources and is not approval. Personalized priorities do not change the published overall score.
- Review and source-check dates describe the research, not this deployment. Fetch current primary sources when the decision requires current verification.
- Filters are evidence thresholds, not guarantees of permission, privacy or ethical conduct. Minimum scores exclude unscored findings. Explain when no entries meet the user's constraints rather than silently relaxing them.
- Use task to select capability-specific findings. Local capability alone does not mean every configuration stays on the device. Compatibility between an app, model and host needs separate evidence.
- Cite the canonical assessment URL and relevant primary sources. Report recorded dates as research dates; do not imply that this lookup reverified the sources.

The service supplies research only. It does not authorize purchases, configuration changes or messages.