Patent Pending: US 63/948,692

The Missing Layer in Autonomous AI

Human Supervision Protocol (HSP) — the world's first framework for mandatory, non-bypassable human oversight in AI agent interactions. EU AI Act compliant by design.

Schedule Consultation Read the Research
€35M
Max EU AI Act Penalty
7%
Of Global Revenue at Risk
2026
Full Compliance Deadline
📋
US Patent Pending Application 63/948,692
📄
Published Research DOI: 10.5281/zenodo.18091551
🇪🇺
EU AI Act Aligned Articles 9, 12, 13, 14
🔒
Fail-Closed Architecture Non-Bypassable by Design
The Problem

AI Agents Are Talking to Each Other.
Who's Watching?

As MCP, A2A, and ANP protocols enable autonomous agent-to-agent communication, a critical question remains unanswered: How do humans stay in control?

🔄

Infinite Loops

AI agents negotiating endlessly without human intervention, consuming resources and making unauthorized commitments.

⚡

Flash Events

Rapid cascading actions between agents—like the 2010 Flash Crash—causing trillion-dollar consequences in minutes.

🚫

Bypassable Oversight

Current "human-in-the-loop" systems can be configured, minimized, or disabled. That's not real oversight.

The Solution

Human Supervision Protocol

A revolutionary framework that makes human oversight mandatory at the architectural level—impossible to bypass, disable, or configure away.

1

Mandatory Approval

Human approval required for all actions above baseline risk. No exceptions. No bypass mechanisms. No "emergency override."

2

Fail-Closed Default

Any timeout, error, or absence of explicit approval results in action DENIAL. Never autonomous execution. Ever.

3

3-Cycle Anti-Loop

AI-to-AI negotiations limited to 3 cycles before mandatory human intervention. Prevents runaway agent interactions.

4

4-Tier Risk Classification

LOW, MEDIUM, HIGH, CRITICAL — each with specific approval requirements, timeout periods, and escalation paths.

Regulatory Ready

EU AI Act Compliance Built In

HSP directly maps to EU AI Act requirements for high-risk AI systems. Stop retrofitting compliance—build it in from day one.

EU AI Act
Requirement
HSP Solution
Article 9
Risk Management Systems
4-tier risk classification with automated controls
Article 12
Record-keeping & Traceability
Blockchain-based immutable audit trails
Article 13
Transparency & Documentation
Complete visibility of decisions and approvals
Article 14
Human Oversight
Mandatory, non-bypassable human approval architecture
Services

How We Can Help

From assessment to implementation, we guide you through building AI systems that are safe, compliant, and trustworthy.

AI Governance Assessment

Custom

Comprehensive review of your AI systems against EU AI Act requirements

  • Risk classification audit
  • Compliance gap analysis
  • Oversight mechanism review
  • Remediation roadmap
  • Executive summary report
Get Quote

HSP Implementation

Custom

Full implementation of Human Supervision Protocol in your AI infrastructure

  • Architecture design
  • Risk tier configuration
  • Approval workflow setup
  • Audit trail integration
  • Anti-loop mechanism deployment
  • Team training
Get Quote

Advisory Retainer

Custom

Ongoing strategic guidance on AI governance and regulatory compliance

  • Monthly strategy sessions
  • Regulatory monitoring
  • Policy development
  • Board presentations
  • Incident response support
Get Quote
About

Jaqueline Martins de Jesus

Independent researcher and entrepreneur focused on AI safety and governance. Inventor of the Human Supervision Protocol (HSP), the first framework to make human oversight architecturally mandatory in autonomous AI systems.

With a background in technology entrepreneurship and a deep commitment to responsible AI development, I help organizations navigate the complex landscape of AI regulation while building systems that are safe, transparent, and trustworthy.

🔒 Patent Pending: US 63/948,692
📄 Published Researcher (Zenodo)
🇪🇺 EU AI Act Specialist
🇬🇧 Based in London, UK

Ready to Get Compliant?

The EU AI Act deadline is approaching. Don't wait until it's too late. Let's discuss how HSP can protect your organization.

Schedule a Call Connect on LinkedIn