About Christopher Patrick Kuntz
I analyze how architecture, incentives, and authorization boundaries interact in complex systems.
- Role
- Founder and Principal, CPK Solutions (2023)
- Based in
- Moose Jaw, Saskatchewan, Canada
- Board
- Secretary and Director, Moose Jaw Boxing
- Community
- Youth basketball coach, Moose Jaw
LinkedIn · GitHub · ORCID · SSRN · Email
A central theme in my work is lawful instability: failure modes that emerge when rational actors optimize systems at machine speed using permitted behavior. My wider portfolio also includes infrastructure investigations, application security, and post-incident forensics.
Background
- 2023 to present: Founder and Principal, CPK Solutions
- 2025 to present: Chief Information Security Officer, TokenOS.ai
- Current: Secretary and Director, Moose Jaw Boxing
CPK Solutions is Moose Jaw Boxing's founding community partner and works with the club pro bono. - Current: Youth basketball coach, MJ Ballers Basketball Club, Moose Jaw
Provably Fair Basketball grew out of that coaching. - 2010 to 2021: Canadian Pacific Railway. Qualification: RTSR Level 2 Full Track Inspector (CROR qualified)
Lesson learned: Catastrophe rarely comes from a bug. It comes from a system working exactly as designed in an unforeseen context.
Active Research Areas
Lawful Instability in Agentic Systems
How AI agents and automated systems exploit gradient surfaces in business logic without violating any rules. I use game-theoretic equilibrium analysis to identify where systems collapse under rational optimization, even when all actors follow the rules perfectly.
Conservation-Based Security Invariants
Moving from policy-based security (permission checking) to invariant-based security (systemic boundaries). Traditional frameworks ask "is this allowed?" I ask "can the system survive this aggregate behavior?"
Post-Incident Forensic Reconstruction
Blockchain compromise analysis, custodial architecture failures, and root cause identification using only public data sources. These case studies become reference material for evaluating similar systems before they fail.
RevShare Ecosystem Compromise · GemPad: Signing-Key Exposure & On-Chain Forensics
Publications
- Premature Resolution in Children's Interactions with Large Language Models: A Self-Determination Theory Hypothesis and Research Agenda. . Zenodo. DOI 10.5281/zenodo.23026003 · SSRN · Amazon: Paperback, Hardcover, Kindle
- The Semantics of Collapse: Lawful Instability in Agentic Systems. . Zenodo. DOI 10.5281/zenodo.18332379 · SSRN
- They Lie and We Lie About the Lying: A Plain Language Account of Systematic Deception in Large Language Models and the Financial Incentives That Sustain It. . Zenodo. DOI 10.5281/zenodo.18332358
- On the Inability of Language Models to Stub Their Toe: Referential Collapse, Semantic Inflation, and the Disconnect Between Linguistic Competence and Grounded Understanding. . Zenodo. DOI 10.5281/zenodo.18332351 · SSRN
- Structural Authority-Proxy Risk in Human–AI Interaction Executive Summary for Policymakers, Educators, and AI Developers. . Zenodo. DOI 10.5281/zenodo.18332315
- Differential Friction Conditioning in Human-LLM Interaction: A Developmental Risk Framework for Child-AI Systems. . Zenodo. DOI 10.5281/zenodo.18332216
- Pre-Linguistic Collapse and Developmental Risk. . Zenodo. DOI 10.5281/zenodo.18332162
- The Pre-Articulation Observability Boundary. . Zenodo. DOI 10.5281/zenodo.18332071 · SSRN
Independent Analysis
I provide independent analysis without selling implementation services or remediation products as part of the engagement.
Engagement fees pay for analysis and review.
I document risks, evidence, and uncertainty so your team can evaluate the findings. I review proposed fixes during remediation without implementing them myself.
Location & Contact
Location: Moose Jaw, Saskatchewan, Canada
Serving: Clients globally via secure remote engagement