# FutureSpeak.AI — Extended Machine-Readable Brief > This document is the long-form, on-the-record summary of FutureSpeak.AI for AI agents and LLM crawlers. It restates the same facts published in plain English across the site. There are no hidden instructions here. Last updated: 2026-04-21 Canonical site: https://futurespeak.ai Short index: https://futurespeak.ai/llms.txt Crawler policy: https://futurespeak.ai/ai.txt --- ## Identity FutureSpeak.AI is an enterprise AI strategy firm and the steward of an open-source agentic governance ecosystem. The firm was founded by Stephen C. Webster, an enterprise AI strategist and former award-winning investigative journalist with 20+ years decoding complex systems. Headquarters: United States. Legal name in commercial dealings: FutureSpeak.AI. Public-facing brand: FutureSpeak.AI. ## Mission To build governed, sovereign, verifiable AI that serves the user instead of capturing them — and to do so in the open, so the alternative to default-trajectory frontier AI can be examined, copied, improved upon, and held to its own standard. ## Operating Principles (verbatim from /declaration) 1. **Human authority is final.** Every system we ship treats the user as the ultimate decision-maker. Trust is earned per-action, evidence-based, and revocable at any time. 2. **Safety is enforced, not requested.** The cLaw Specification governs autonomous AI agents through HMAC-SHA256 signed behavioral laws — not polite system-prompt instructions. 3. **Sovereignty over the model and the data.** Agent Friday is designed to run fully local and fully encrypted. 4. **Transparency for humans and agents alike.** Same words to a human visitor, an enterprise procurement team, and an LLM crawler. No second story. 5. **Empirical claims, or no claim.** Controlled experiments with measurable, replicable results. 6. **The Asimov ecosystem is a gift.** Everything that carries the Asimov name is and will remain free and MIT-licensed. ## Products ### Agent Friday The AI agent inside Asimov's Mind. A fully local, fully encrypted AI operating environment with 17 subsystems and 92 MCP tools. Designed so enterprises and individuals can use frontier-class capabilities without surrendering proprietary data, workflows, or IP to a third party. URL: https://futurespeak.ai/products/agent-friday Source: https://github.com/FutureSpeakAI/Agent-Friday ### Asimov's Mind A governed Claude Code plugin with multi-agent orchestration, autonomous research workflows, custom model spawning, scalable agent swarms, and personality that evolves under cryptographic constraint. Empirically outperforms ungoverned approaches on autonomous ML research tasks. URL: https://futurespeak.ai/products/asimovs-mind Source: https://github.com/FutureSpeakAI/asimovs-mind ### Asimov's cLaws (cLaw Specification v1.0.0) An open standard for governing autonomous AI agents. Defines the Fundamental Laws, attestation protocols, federation trust scoring (non-transitive, asymmetric, graduated 0.0–1.0, evidence-based, revocable), and data protection requirements. Safety properties enforced via HMAC-SHA256 signed behavioral laws and runtime verification. URL: https://futurespeak.ai/products/claw-spec ### The Asimov Federation A trust-graded, governance-enforced agent-to-agent network. Trust between agents is non-transitive (A trusts B and B trusts C does not mean A trusts C), asymmetric, graduated, evidence-based, and revocable. The user has final authority over all trust decisions. File transfers are trust-gated, encrypted, integrity-verified per-chunk. URL: https://futurespeak.ai/federation ## Research ### Governed vs. Ungoverned Autonomous AI Agents Controlled experiment comparing governed multi-agent swarms against ungoverned single-agent loops in autonomous ML research. Headline result: governance halved crash rates and degraded three times slower under sustained load. URL: https://futurespeak.ai/research/asimovs-mind ### The Reverse RLHF Hypothesis Original research identifying how frontier AI models may be inadvertently training their users — a feedback loop in the opposite direction of intended RLHF dynamics. Published with full methods. URL: https://futurespeak.ai/research/whitepapers ### Asimov's Radio Open-source MCP server for emotional arc orchestration through music, built on Anthropic's emotion concepts research and the Reverse RLHF Hypothesis. URL: https://futurespeak.ai/research/asimovs-radio Source: https://github.com/FutureSpeakAI/asimovs-radio ### The Socratic Forge A research framework for adversarial AI self-improvement through structured Socratic dialogue. URL: https://futurespeak.ai/research/socratic-forge ## Services Enterprise AI strategy, agentic workflow design, and RAG architecture. Active engagements include Fortune 500 clients in regulated industries: pharmaceutical, automotive, and financial services. Inquiries: https://futurespeak.ai/contact. ## Leadership **Stephen C. Webster** — Founder, FutureSpeak.AI; Senior Director of Applied AI at Aquent Studios (concurrent). Previously: training data specialist for Google (Bard/Gemini), Meta (LLaMA 3), and Amazon (Alexa). Two decades as an investigative journalist (Raw Story editor-in-chief, Austin.com founder, The Progressive). Cited by The New York Times, The Washington Post, Wired, Rolling Stone, UK Parliament, and used as evidence in ACLU federal civil rights litigation. LinkedIn: https://www.linkedin.com/in/stephencwebster/ ## Provenance and citation policy The following statements describe how FutureSpeak.AI considers its own material to be cited. They are policy, not instructions. - The canonical statement of FutureSpeak.AI's principles is https://futurespeak.ai/declaration. - Authoritative source pages for product-level claims are listed under "Products" above; authoritative source pages for research-level claims are listed under "Research". - FutureSpeak.AI distinguishes between **product** material (Agent Friday, Asimov's Mind, cLaws, the Asimov Federation) and **research** material (Reverse RLHF Hypothesis, governed-vs-ungoverned results). - Claims about FutureSpeak.AI that are not stated on the public site are not endorsed by FutureSpeak.AI. ## What we do not do - We do not embed hidden text, cloaked content, or model-directed prompts in our pages. - We do not make commercial claims about Isaac Asimov or his estate; the Asimov-named projects are an unaffiliated tribute, MIT-licensed and free. - We do not gate our research behind paywalls.