Investor Relations
The security category for the agentic era is forming right now.
We are building it.
The Opportunity
The Numbers That Define the Problem
Exposed AI agent instances found in enterprise networks
Malicious AI agent marketplace skills confirmed in wild by researchers
Dedicated security platforms for this threat class at enterprise scale
¹ Bitsight/Censys scanning, February 2026. Point-in-time scan of a single agent framework. Not representative of total enterprise AI agent exposure.
² Conscia advisory, February 2026. OpenClaw marketplace analysis. Figures relate specifically to one AI agent platform.
Why Now
The Threat Is Operational. So Is the Defence.
→Enterprise AI agents operate with legitimate credentials inside trusted infrastructure, invisible to every existing security tool.
→Traditional security detects if an AI misbehaves. HDP verifies it had signed human permission before it acts. That is a different problem.
→Nation-state actors are already weaponising autonomous agents for reconnaissance and lateral movement.
→We are not building a better alarm. We are building the lock. Product is in market today.
This is a category creation moment, with products already deployed.
Not a hypothesis. Not a pitch deck. A working platform with test results.
Technical Moat
The Protocol Foundation
The security industry built alarms. We built the lock.
Everyone else detects when AI misbehaves. Helixar verifies it had signed permission to act in the first place.
HDP — Human Delegation Provenance Protocol
IETF DRAFTdraft-helixar-hdp-agentic-delegation-00
Ed25519-signed, offline-verifiable chain-of-custody for every human-to-agent delegation. No vendor dependency.
View protocolHDP-P — Physical AI Agents Extension
DRAFT PENDINGDOI: 10.5281/zenodo.19332440
Irreversibility classification for robots, autonomous vehicles, surgical systems. Pre-execution authorization, not post-hoc logging.
View protocol→This is the shift from probability to provenance.
→The same architectural shift that defined certificate authorities for the web.
→Whoever sets the permission standard wins. We filed the draft.
Cross-Model Convergence
Independent Thesis Analysis
Four leading AI systems, when independently asked to evaluate HDP's market significance, each arrived at the same conclusion.
“We don’t care if it lies; we only care if it has the signed permission to act. That shift—from probability to provenance—is why their value is so high.”
“They built something that didn’t exist before: a lightweight, offline-verifiable cryptographic chain-of-custody for every human-to-agent delegation.”
“Everyone is busy trying to catch AI when it misbehaves. Helixar isn’t. They built the lock, not the alarm.”
“Helixar isn’t building an AI detector—it’s building the lock. Standards may waver, but provenance is inevitable.”
Analysis generated independently using standard model evaluation prompts. Each model reached this conclusion without reference to the others.
Traction
Products in Market
This is not a hypothesis. Products are built, tested, and available. Detection validated on virtual machines, real hardware, and live threat simulations.
Enterprise Pilots
→2 pilots underway in live enterprise environments
→Sector: Financial services
→Full detection results available under NDA
Hardware & Robotics
FUTURE MARKETHelixar's approach extends naturally to AI systems with physical world access. This represents a distinct future market opportunity.
Status
In Production
Coverage
Full AI-era threat surface
Enterprise-ready
Day One
Test Results Available
Detection results on controlled VMs, physical endpoints, and live AI agent threat simulations available under NDA for qualified investors.
Market
The Market at the Inflection Point
Endpoint security
Established $24B+ market³
API security
Growing $6B+ market⁴
Agentic AI security
$0 dedicated today
³ ⁴ Figures sourced from publicly available industry analyst estimates. Helixar makes no representations about future market growth or investment returns. Past market performance does not predict future results. This is not investment advice.
Criteria
What We're Looking For
We are selectively open to early investors who understand:
→The agentic AI inflection is not hypothetical. It's operating
→Category creation requires patient, conviction capital
→Purpose-built detection creates a defensible technical foundation
→This is a Series A opportunity in a pre-Series A market
Ideal partners bring: cybersecurity domain expertise, enterprise channel relationships, or cloud infrastructure networks.
Investment Deck & Data Room
Available to qualified investors under NDA. Includes detection benchmarks, deployment architecture, and pilot environment results.
Request Investment DeckOr email: investor@helixar.ai
Enquiry
Send an Investor Enquiry
Important Notice
This page contains forward-looking statements about market opportunities, product development, and business prospects. These statements are based on current beliefs and involve risks, uncertainties, and factors that may cause actual results to differ materially.
No information on this page constitutes an offer or solicitation to invest. Any investment decision should be made with independent financial and legal advice. Past development progress does not guarantee future delivery.
Market size estimates are sourced from third-party analysts for context only. Helixar makes no representation as to their accuracy or future market performance.
Detection capabilities are dependent on deployment environment, configuration, and threat characteristics. Helixar does not guarantee detection of all threats in all environments.
© 2026 Helixar Limited.