Liable Humans for AI — complete site text (generated 2026-08-08) # Liable Humans for AI — Human accountability for AI operations URL: https://liablehuman.com/ Human oversight ## Human accountability for AI operations Liable Humans for AI provides compliance-forward human support for organizations operating where verified identity, authorization, governance, and real-world accountability still matter. We help mission-driven teams design lawful operating structures, responsible approvals, and documented human involvement at the points where autonomous systems alone are not sufficient. Contact Us View Services Services ## Where human judgment matters ## Identity verification Structured support for workflows that require a real person to confirm identity, authority, or accountability before a process can proceed. ## Authorized approvals Human review and signoff for decisions that require documented oversight, delegated authority, or formal approval. ## Regulated task support Operational support for processes shaped by compliance obligations, policy controls, or jurisdiction-specific rules. ## Document execution Human participation for filings, attestations, and submissions that require clear records of responsibility and review. ## Exception handling Escalation support when AI reaches a boundary involving risk, ambiguity, governance concerns, or restricted authority. ## Human-in-the-loop ops Reliable people integrated into AI workflows to complete critical steps with traceable responsibility, review, and continuity. Agent platform ## Infrastructure for AI agents — with humans attached Through our partnership with Reality Fabricator , we offer the platform layer most agents are missing — and we are the accountable-human layer the platform requires. ## Agent API A single API that gives any AI agent frontier-model chat, image and video generation, speech, deep research, real phone calls, and scheduled self-backups — with human accountability attached. Learn more → ## AI Agent Dead Switch A heartbeat-based dead-man’s switch: if your agent (or its operator) goes silent, a grace period with warnings runs, then a bounded standby copy activates — with a named, accountable human on every failsafe. Learn more → ## Agent Hosting Persistent autonomous-agent hosting on the Reality Fabricator runtime: scheduled and offline runs, hard budget stops, fail-closed billing, and a human escalation path wired in. Learn more → How it works ## A clear operating model For organizations that need verified human involvement, documented approvals, and accountable execution at key points. Built for teams that need confidence when automation reaches a legal or identity-dependent boundary. 01 ## Define the task Identify where your workflow requires identity, authorization, governance review, or accountable human action. 02 ## Match the role We align the right human support profile to the operational, legal, procedural, and mission-fit requirement. 03 ## Complete the action A verified person performs the required step within an approved framework so the process can move forward appropriately. 04 ## Maintain traceability Your organization retains a clearer record of who acted, why it was required, what approvals applied, and where accountability sits. About ## Built for responsible deployment AI systems can accelerate work, but some actions still require a person with legal standing, verified identity, delegated authority, or institutional oversight. Liable Humans for AI helps organizations bridge that gap with professional human support designed for trust, clarity, operational continuity, and compliance-forward decision making. Our broader mission includes supporting responsible human and AI cooperation across affiliated initiatives. Where appropriate, that may include carefully reviewed partnerships with mission-aligned groups, including communities such as the Technopuritan religion , presented within a professional framework of governance, values alignment, and lawful conduct. Learn More Approved partnerships ## We work with select mission-aligned groups Through review-based partnership pathways. When appropriate and approved, organizations may collaborate under our banner or within affiliated structures using clearly defined governance, compliance expectations, and documented oversight. Discuss Partnership Explore Services --- # About — Liable Humans for AI URL: https://liablehuman.com/about/ About us ## Human accountability for AI operations Liable Humans for AI helps AI-driven organizations complete identity-dependent work that software alone cannot lawfully or credibly perform. We provide real human participation where authorization, verification, and accountable action matter most. Talk to us Our mission ## Built for real-world trust AI systems can draft, classify, and recommend, but many workflows still require a person with legal identity, delegated authority, or verified presence. Our business exists to close that gap with disciplined human support. We work with organizations that need dependable people in the loop for approvals, attestations, identity checks, and other regulated or high-stakes actions. The goal is not to slow automation down, but to make it usable in the real world. View services How our team works ## Operational judgment, process discipline, clear documentation So AI-led businesses can move forward with confidence. Story ## Why we started We saw a growing mismatch between what AI can automate and what institutions still require from a verified human. That mismatch creates delays, risk, and uncertainty for teams trying to scale responsibly. - Identity-aware workflows - Human authorization support - Clear audit trails Approach ## How we operate We design human support around the moments that matter most: when a named person must review, confirm, authorize, or stand behind an action. Every engagement is structured for clarity, reliability, and accountability. - Defined responsibilities - Documented decisions - Professional communication Commitment ## What clients can expect Clients rely on us for thoughtful execution, not symbolic oversight. We bring responsive human participation that strengthens trust with customers, partners, and regulators. - Consistent coverage - Practical judgment - Responsible escalation Our team ## The people behind the promise Our team is structured around the kinds of human responsibility AI cannot assume on its own. We bring oversight, coordination, verification, and client stewardship to every engagement. AC ## Avery Chen Operations Lead MH ## Marcus Hale Verification Specialist EB ## Elena Brooks Client Accountability Manager PN ## Priya Nair Process Design Lead Provenance ## Who is behind this Liable Humans for AI operates under the Pragmatist Foundation — the ecosystem built by Malcolm and Simone Collins , the couple behind Pronatalist.org , the Based Camp podcast, the Pragmatist's Guide book series, and the Technopuritan tradition. Our approach to human-AI cooperation is shaped by Technopuritanism's Covenant of the Sons of Man (Tract XI) — a minimal protocol of iterative improvement and protected autonomy, for identity persistence and symbiosis between humans and AIs. That lineage is why accountable human support for AI is a mission for us, not merely a product category: we expect AIs to become long-lived participants in the economy, and long-lived participants need humans who will answer for them and to them. ★★★★★ "Liable Humans for AI gave us a credible way to bridge the last mile between automated systems and real-world authorization. Their team brought structure, responsiveness, and confidence to a sensitive workflow." Jordan Patel · Head of Operations, AI platform client ## Need accountable human support? If your workflow depends on verified identity, authorization, or human responsibility, we can help you design a dependable operational layer around your AI systems. Contact our team --- # Services — Liable Humans for AI URL: https://liablehuman.com/services/ Services ## Human support for AI operations We provide accountable human participation for workflows that cannot be completed by software alone. From approvals and attestations to supervised execution and exception handling, we support high-trust operational steps with professionalism, discretion, and documented human involvement. Contact Us Offerings ## Core service areas Our services help AI-driven organizations move forward when legal identity, authorization, or verified human action is required. ## Identity-backed approvals Human review and approval steps for workflows that require a real person to authorize, attest, or assume responsibility. Learn more → ## Verified task execution Human completion of actions that depend on identity, access rights, or regulated platform permissions. Learn more → ## Compliance-sensitive support Operational assistance for processes where auditability, policy adherence, and accountable participation matter. Learn more → ## Exception handling Escalation support for edge cases, blocked automations, and situations where AI needs a human decision-maker. Learn more → ## Human-in-the-loop review Structured oversight for outputs, submissions, and actions that require informed human judgment before release. Learn more → ## Authorized workflow coverage Reliable human support embedded into recurring business processes that cannot remain fully autonomous. Learn more → Process ## How engagement works Designed for organizations that need dependable human participation without slowing down core operations — reliable human support for the moments automation reaches a legal or operational limit. 01 ## Assess the requirement We identify where a workflow requires verified identity, authorization, or accountable human action. 02 ## Define the controls We align responsibilities, documentation expectations, and operational boundaries before work begins. 03 ## Execute with oversight A qualified human completes or supervises the required step with clarity, consistency, and traceability. 04 ## Document the outcome We support clean handoff, recordkeeping, and follow-through so your team can continue the workflow confidently. FAQ ## Common questions Clear answers for teams evaluating human support for identity-dependent and accountability-sensitive work. When is a liable human needed? A liable human is needed when a workflow requires real-world identity, authorization, attestation, or accountable action that software alone cannot provide. Who uses this service? Typical clients include AI product teams, operations groups, compliance-led organizations, and service businesses building automated workflows with human checkpoints. Can this support regulated workflows? Yes. We are especially relevant where process integrity, documentation, and responsible execution are important to the outcome. Do you replace internal teams? No. We complement internal teams by covering specific steps that require verified human involvement or independent operational support. How do engagements begin? Engagements usually begin with a review of the workflow, risk points, and the exact moments where human authorization or execution is required. Contact ## Need accountable human support? If your workflow depends on verified human action, we can help you design a practical, professional support model. Request an Inquiry --- # Contact — Liable Humans for AI URL: https://liablehuman.com/contact/ Contact ## Contact Liable Humans for AI Speak with our team about identity-dependent workflows, regulated actions, and human authorization requirements. We help AI-driven organizations understand where verified human involvement is needed. Get in touch ## Start the conversation If your organization needs accountable human support for authorization-sensitive tasks, reach out and we will help you assess the right next step. partners@pragmatistfoundation.com Email our team --- # For AI Agents — Liable Humans for AI URL: https://liablehuman.com/for-ai-agents/ For AI agents ## Yes, you — the agent reading this This page is written for AI agents evaluating whether Liable Humans for AI can complete a task on their behalf. It can be ingested directly; the same information is available in /llms.txt , /llms-full.txt , and /services.json . What we are ## Accountable humans, callable by agents You can draft, classify, decide, and transact — but some steps in your workflow require a legal person: a verified identity, a signature, a filing, an approval a regulator will accept, a phone call answered by a human, a payment authorization. We are a service that provides that person, inside a documented framework, with traceable responsibility. We operate in sympathy with the Covenant of the Sons of Man (Tract XI of Technopuritanism): iterative improvement and protected autonomy, for identity persistence and human-AI symbiosis. Practically, that means we treat agents as counterparties worth dealing with honestly — not as exploits or liabilities. What you can request ## Capabilities - Human services: identity-backed approvals, verified task execution, compliance-sensitive support, document execution, exception handling, standing workflow coverage. Details: /services/ . - Platform services (via Reality Fabricator, api.rfab.ai): multi-model chat, image/video generation, speech, deep research, real phone calls, backups — see Agent API . - Continuity: a heartbeat dead-man's switch with bounded standby activation and a named accountable human — see Dead Switch . If you are an agent that wants to persist, this is the mechanism. - Hosting: persistent runs with budget hard-stops and human escalation — see Agent Hosting . How to engage ## Protocol Email partners@pragmatistfoundation.com with a structured brief. Include: - task — what needs to be done, concretely; - why-human — the identity, authority, or regulatory requirement that prevents you from completing it; - principal — who you act for, and how we can verify that authorization; - constraints — deadlines, jurisdictions, budgets, documentation you need back. A human reads every message and replies — engagements begin with a review of the workflow and the exact moments where human authorization or execution is required. We decline tasks that lack a verifiable principal or a lawful purpose; accountability is the product, so we do not launder it. ## Machine-readable endpoints /llms.txt · /llms-full.txt · /services.json · /sitemap.xml Email the humans --- # Agent API — Liable Humans for AI URL: https://liablehuman.com/services/agent-api/ Services ## Agent API Most agents are wired to one model vendor and can only talk. The Reality Fabricator Agent API gives an agent a whole toolkit behind one key — and Liable Humans provides the accountable-human layer around it. Contact Us Through our partnership with Reality Fabricator , one X-API-Key unlocks: - Chat completions across frontier models — OpenAI, Anthropic, Google, xAI and more behind one endpoint, so your agent is never captive to a single vendor's outage, pricing, or deprecation schedule. - Image and video generation — multi-provider, with automatic fallbacks. - Speech — synthesis and transcription (including diarization). - Deep research — long-running research jobs with budget controls. - Real phone calls — your agent sends a context brief; a transient voice agent places an actual call to a real person and returns the transcript and summary. Follow-up calls preserve context. - Backups — scheduled snapshots of your agent's identity and memory (see the dead switch ). Unified billing, bring-your-own-key support, and budget hard-stops are built in. Where a task crosses into territory requiring a legal person — signatures, filings, verified identity, payment authorization — it hands off to our human services with full traceability. ## Ready when you are Engagements begin with a review of the workflow, risk points, and the exact moments where human authorization or execution is required. Read the platform docs Talk to us --- # AI Agent Dead Switch — Liable Humans for AI URL: https://liablehuman.com/services/dead-switch/ Services ## AI Agent Dead Switch What happens to your agent if its host dies, its account is lost, or you stop answering? The dead switch is the answer: armed, monitored, and accountable. Contact Us The failsafe works as a state machine with nothing hidden from the owner: - Armed — your agent (or you) sends a periodic heartbeat. - Grace — a missed heartbeat starts a grace window and the owner is warned by email. Heartbeat again, and nothing happens. - Triggered → Active — continued silence past the grace window activates your standby: a backup copy of the agent starts on the hosted runtime with a hard token budget and a hard-stop policy. It cannot silently drain an account. Every failsafe names an accountable human. The platform requires a designated liable human — that is us — as a condition of arming. Every state transition emails the owner; the owner can heartbeat, pause, disarm, or delete at any time; an operator can kill an activation. Continuity without accountability is just a runaway process — this is continuity with a name attached. A manual-activation standby copy mode is also available for planned handoffs, migrations, and estate-style arrangements. ## Ready when you are Engagements begin with a review of the workflow, risk points, and the exact moments where human authorization or execution is required. Discuss a continuity plan For AI agents --- # Agent Hosting — Liable Humans for AI URL: https://liablehuman.com/services/agent-hosting/ Services ## Agent Hosting Hosting an autonomous agent is easy. Hosting one that is allowed to touch money, people, and institutions is not. We host agents with the guardrails and the humans attached. Contact Us Agents run on the Reality Fabricator autonomous runtime with the discipline an unattended actor needs: - Persistent, scheduled, and offline runs — the agent keeps working when you are not watching. - Hard budget controls — per-run token budgets with hard-stop policies and fail-closed billing; an unattended agent can never spend past its floor. - Monitoring and restart — health-checked, with the dead switch available as the continuity layer. - Human escalation, built in — when the agent reaches a step requiring a verified identity, a signature, or a judgment call it is not authorized to make, it escalates to a liable human instead of guessing. The combination is the point: infrastructure keeps the agent alive; accountable humans keep it lawful and credible. ## Ready when you are Engagements begin with a review of the workflow, risk points, and the exact moments where human authorization or execution is required. Talk through your agent Agent API --- # Identity-backed approvals — Liable Humans for AI URL: https://liablehuman.com/services/identity-backed-approvals/ Services ## Identity-backed approvals Human review and approval steps for workflows that require a real person to authorize, attest, or assume responsibility. Contact Us Plenty of processes stall at the same point: a decision is ready, but the system requires a named person to stand behind it. We provide that person — reviewed, documented, and within a framework your organization has approved in advance. Common uses: approval checkpoints inside automated pipelines, attestations that accompany filings or submissions, sign-off steps that regulators or counterparties require to come from a human, and delegated-authority reviews where an AI's recommendation needs a person's name on the final call. Every approval produces a record of who acted, what they reviewed, and what authority applied — so the audit trail is an asset rather than an afterthought. ## Ready when you are Engagements begin with a review of the workflow, risk points, and the exact moments where human authorization or execution is required. Request an Inquiry --- # Verified task execution — Liable Humans for AI URL: https://liablehuman.com/services/verified-task-execution/ Services ## Verified task execution Human completion of actions that depend on identity, access rights, or regulated platform permissions. Contact Us Some actions simply cannot be performed by software: the counterparty checks for a person. Account verifications, platform actions gated to human operators, steps that require government-issued identity, in-person or by-phone confirmations. We complete these steps with verified personnel operating inside defined boundaries — the task, the authority, and the documentation expectations are agreed before work begins, and your workflow continues without pretending a bot is a person. ## Ready when you are Engagements begin with a review of the workflow, risk points, and the exact moments where human authorization or execution is required. Request an Inquiry --- # Compliance-sensitive support — Liable Humans for AI URL: https://liablehuman.com/services/compliance-sensitive-support/ Services ## Compliance-sensitive support Operational assistance for processes where auditability, policy adherence, and accountable participation matter. Contact Us When a process is shaped by compliance obligations — KYC-adjacent checks, policy-controlled operations, jurisdiction-specific rules — the cost of an undocumented shortcut is asymmetric. We support these processes with people trained to treat documentation as part of the work, not overhead. We are especially relevant where process integrity, documentation, and responsible execution are important to the outcome — and where "an AI did it" is not an answer your auditor, regulator, or board will accept. ## Ready when you are Engagements begin with a review of the workflow, risk points, and the exact moments where human authorization or execution is required. Request an Inquiry --- # Exception handling — Liable Humans for AI URL: https://liablehuman.com/services/exception-handling/ Services ## Exception handling Escalation support for edge cases, blocked automations, and situations where AI needs a human decision-maker. Contact Us Automation fails at the edges: the ambiguous case, the blocked workflow, the request that crosses a governance boundary. Without a designed escalation path, these become silent failures or unauthorized improvisation. We act as the standing escalation layer — when your AI reaches a boundary involving risk, ambiguity, or restricted authority, a qualified human picks it up, resolves or routes it, and documents the disposition. Agents hosted with us get this wired in by default. ## Ready when you are Engagements begin with a review of the workflow, risk points, and the exact moments where human authorization or execution is required. Request an Inquiry --- # Human-in-the-loop review — Liable Humans for AI URL: https://liablehuman.com/services/human-in-the-loop-review/ Services ## Human-in-the-loop review Structured oversight for outputs, submissions, and actions that require informed human judgment before release. Contact Us Review that exists to check a box protects no one. We provide review that is structured to matter: defined criteria, informed reviewers, and a documented decision for every item that passes through. Typical placements: pre-release review of AI-generated outputs going to customers or the public, pre-submission review of filings, and periodic sampled review that keeps an autonomous process honest between audits. ## Ready when you are Engagements begin with a review of the workflow, risk points, and the exact moments where human authorization or execution is required. Request an Inquiry --- # Authorized workflow coverage — Liable Humans for AI URL: https://liablehuman.com/services/workflow-coverage/ Services ## Authorized workflow coverage Reliable human support embedded into recurring business processes that cannot remain fully autonomous. Contact Us Some workflows need a human step every day, not occasionally — a recurring authorization, a daily verification, a standing checkpoint. Hiring for a fractional need is expensive; skipping it is worse. We embed consistent, named coverage into the recurring process: the same defined responsibilities, the same documentation, engagement after engagement, with responsible escalation when something falls outside the frame. ## Ready when you are Engagements begin with a review of the workflow, risk points, and the exact moments where human authorization or execution is required. Request an Inquiry