AI & Cybersecurity Strategist
AI and Cybersecurity Don't Collide. They Converge.
I'm Joshua Russell, a cybersecurity engineer and AI strategist based in Sydney. I explore the critical intersection where artificial intelligence meets enterprise security — building tools, writing about emerging risks, and helping organisations navigate what's next.
Areas of Focus
What I Work On
Deep expertise at the intersection of artificial intelligence and security, with a practitioner's mindset.
AI Strategy & Security
The risks, opportunities, and architecture of AI systems. Understanding how LLMs, agents, and AI infrastructure introduce new attack surfaces — and how to defend them.
Enterprise Cybersecurity
Threat landscapes, SIEM architecture, cloud security posture, and identity. Designing defense-in-depth strategies for modern hybrid infrastructure.
AI Communication
Deep intuition for how LLMs process and respond. Closing the gap between intent and output to make AI a reliable engineering tool, not a slot machine. Years of real-world use building production systems.
Philosophy
AI Isn't Magic. It's a Language.
Most people treat AI like a search engine — ask a question, hope for a useful answer. I see it differently. AI is like meeting a new alien race: intelligent, capable, but with its own operating logic. The gap between what you ask and what you get isn't a bug. It's a translation problem.
I've spent years learning the language. I know how to close the intent-output gap, how to make the non-deterministic deterministic enough, and how to build production systems on top of AI that others say can't be automated yet.
Read My ApproachPerspectives
Featured Writing
Analysis and insights on AI infrastructure, security implications, and the technology shaping our industry.
The AI Threat Landscape in 2026: What Defenders Actually Need to Worry About
A high-level, evidence-led briefing on AI-enhanced attacks, agent security, supply-chain risk, data exposure, deepfakes, and what defenders should do now.
Read Article AIGPT-5.6 Is Here: Sol, Terra, and Luna Explained
A concise, fact-checked guide to when GPT-5.6 Sol, Terra, and Luna make sense, what the benchmarks show, and when to raise effort.
Read Article AIHow to Build a Self-Improving Memory System in Claude Code
Wire four config layers — auto memory, instruction files, skills, and hooks — so the agent stops repeating mistakes and improves itself across sessions.
Read ArticlePortfolio
What I've Built
Applied research and tools — because the best insights come from building.
AI Security Lab
Mythos exposure lab, gateway checklists, scorecards, and mini-labs that turn AI security attention into practical reviews.
Run Mythos Exposure LabIran–Israel Intel Dashboard
Real-time conflict intelligence — interactive attack map, D3 actor network graph, live OSINT feeds, threat actor profiles, and incident timeline. Updated daily.
Open DashboardTidyKit.tech
A developer utility platform with practical tools for everyday engineering workflows. JSON/YAML converters, formatters, encoders, and more — built for speed and simplicity.
Visit TidyKitScriptception
This platform — a hub for thought leadership, technical writing, CTF challenges, and developer tools. Built as a portfolio and community resource for security practitioners.
Learn More