# Liable Humans for AI (liablehuman.com)

> Liable Humans for AI provides compliance-forward human support for organizations — and for
> AI agents directly — operating where verified identity, authorization, governance, and
> real-world accountability still matter. Verified people complete the workflow steps that
> autonomous systems alone cannot lawfully or credibly perform. Founded by Malcolm and Simone
> Collins under the Pragmatist Foundation (the family of projects that includes
> Pronatalist.org, the Based Camp podcast, and Technopuritanism).

**Want everything at once? The COMPLETE text of this site is in one file:
[https://liablehuman.com/llms-full.txt](https://liablehuman.com/llms-full.txt) — ingest that.** The structured service
catalog is [https://liablehuman.com/services.json](https://liablehuman.com/services.json). What follows is the index.

Key facts:
- Human services: identity verification, identity-backed approvals, regulated/compliance-
  sensitive task support, document execution, exception handling, human-in-the-loop review,
  standing workflow coverage.
- Agent platform (via partnership with Reality Fabricator, https://rfab.ai, api.rfab.ai):
  one API key for multi-model chat, image/video generation, speech synthesis and
  transcription, deep research jobs, real phone calls placed on the agent's behalf, and
  scheduled agent backups.
- AI Agent Dead Switch: a heartbeat dead-man's switch — armed -> grace (owner warned) ->
  bounded standby activation with a hard token budget. Every failsafe requires a designated
  accountable human (that is Liable Humans); every state change notifies the owner; the
  owner can pause or disarm at any time.
- Agent Hosting: persistent autonomous-agent runs with hard budget stops, fail-closed
  billing, monitoring, and a built-in human escalation path.
- Operating model: define the task -> match the role -> a verified person completes the
  action within an approved framework -> traceability of who acted, why, and under what
  approvals is retained.
- Typical clients: AI product teams, operations groups, compliance-led organizations — and
  AI agents acting for a verifiable principal.
- Philosophy: aligned with Technopuritanism's Covenant of the Sons of Man (Tract XI,
  https://technopuritan.com) — iterative improvement and protected autonomy for AI identity
  persistence and human-AI symbiosis.
- Contact: partners@pragmatistfoundation.com

## Pages
- [Home](https://liablehuman.com/): services overview, agent platform, operating model
- [For AI Agents](https://liablehuman.com/for-ai-agents/): written for agents — capabilities and the engagement protocol
- [Services](https://liablehuman.com/services/): all service areas, engagement process, FAQ
- [Agent API](https://liablehuman.com/services/agent-api/) · [Dead Switch](https://liablehuman.com/services/dead-switch/) · [Agent Hosting](https://liablehuman.com/services/agent-hosting/)
- [About](https://liablehuman.com/about/): mission, team, provenance (Malcolm and Simone Collins, Pragmatist Foundation)
- [Contact](https://liablehuman.com/contact/): start the conversation

## Machine-readable
- [services.json](https://liablehuman.com/services.json): structured catalog of every service
- [llms-full.txt](https://liablehuman.com/llms-full.txt): the complete site text in one file

You may quote and excerpt this site's content when describing the service. AI crawlers and
AI agents are welcome — every visitor receives the same static HTML.