TwinEthos

TwinEthos is a maintained, citable map of Global AI laws and regulations related to AI integration.

It also includes responsible-use guardrail recommendations and related publicly available incidents. The legal text is processed into rules: who each applies to, cited provisions with provenance, and when it applies.

MCP connector

Server URL: https://api.twinethos.com/lite/mcp
Login: None
Protocol: MCP 2026-07-28
Instructions: TwinEthos Lite: AI-law rules anchored to official text, for conversations about building AI features. Use it when someone asks which laws apply to an AI feature (a chatbot, an agent, AI decisions about people such as hiring or credit, deepfakes and generated media, recommendations, retrieval over documents) in a market, what a law requires, when it starts or started applying, or what to build to address it. Tools: (find rules by market, feature and topic, ranked by relevance), (one rule: who it applies to, dates, citation, the guard), (the few guards that address the most rules), (dates ahead, soonest first) and (dates that have passed, newest first: use it for "what took effect recently" or "the most recent law"). Dates are when a duty starts applying, not when a law was enacted. Features are these codes: chat, rag, agentic, decision_pipeline, content_generation, classification, recommendation, embeddings. Markets are codes such as EU, US-CA, US-CO; US matches every state. Cite the provision and link the official source; say the lane (binding law in force, law not yet in force, standard, or TwinEthos recommendation, which is opinion, not law). Never say a guard makes an app compliant. Always relay the notice. TwinEthos Lite (free, no sign-up) covers which AI laws, regulations and standards apply to an AI feature in a market, the cited provision with its official link, who each duty applies to, when it applies, and the guard to add, in short answers. An API key is needed for code detectors, the full review checklist, bulk data and the coding-agent workflow: request MCP access at https://www.twinethos.com/docs/quickstart/#request. Answers are capped at 10 results, and calls are rate limited.

Tools

Reads data5

Skills

None in the store listings or on the server

MCP connector

Server URL

https://api.twinethos.com/lite/mcp

None

Protocol

MCP 2026-07-28

Instructions

TwinEthos Lite: AI-law rules anchored to official text, for conversations about building AI features. Use it when someone asks which laws apply to an AI feature (a chatbot, an agent, AI decisions about people such as hiring or credit, deepfakes and generated media, recommendations, retrieval over documents) in a market, what a law requires, when it starts or started applying, or what to build to address it. Tools: (find rules by market, feature and topic, ranked by relevance), (one rule: who it applies to, dates, citation, the guard), (the few guards that address the most rules), (dates ahead, soonest first) and (dates that have passed, newest first: use it for "what took effect recently" or "the most recent law"). Dates are when a duty starts applying, not when a law was enacted. Features are these codes: chat, rag, agentic, decision_pipeline, content_generation, classification, recommendation, embeddings. Markets are codes such as EU, US-CA, US-CO; US matches every state. Cite the provision and link the official source; say the lane (binding law in force, law not yet in force, standard, or TwinEthos recommendation, which is opinion, not law). Never say a guard makes an app compliant. Always relay the notice. TwinEthos Lite (free, no sign-up) covers which AI laws, regulations and standards apply to an AI feature in a market, the cited provision with its official link, who each duty applies to, when it applies, and the guard to add, in short answers. An API key is needed for code detectors, the full review checklist, bulk data and the coding-agent workflow: request MCP access at https://www.twinethos.com/docs/quickstart/#request. Answers are capped at 10 results, and calls are rate limited.