Walter Niemi

 I care deeply about connection. Family, community, and clarity shape how I approach everything I build and study. That sense of belonging stays at the center of my work, whether I’m working with young people or exploring new frontiers in technology.
 

The AI race unlocked a future for safer cybersecurity systems.
The frontier will belong to those who can best adapt with AI.

A core theme from my research is that many large losses follow the same pattern: high-stakes decisions get made on unauthenticated artifacts, documents, narratives, reputations, and “professional-looking” signals without rigorous provenance checks. In my analysis of high-dollar lending fraud dynamics, critical lessons learned were when organizations skip third-party verification, fail to pressure-test documentation, and rely on thin review layers, the system becomes easy to hack by anyone who can present a coherent story. 

Defense in depth for trust is necessary, including verification architecture, slow-down protocols before disbursement, and explicit trust triggers that force accountability at moments of high risk.

I wrote about insider threat and operational security, specifically, why conventional AI approaches often underperform in real environments. If an analyst pulls 100 records instead of 10, that may be benign or malicious; without context, the system does nothing. When I wrote about Coinbase, the key point was that insiders succeed when they can plausibly and easily hide behind process like “support work,” “tickets,” “customer cases” while the system fails to ever ask the question does the operational story match the underlying access reality?

Cybersecurity must include narrative-aware security controls. The differentiator is building systems that reconcile claims (tickets, approvals, justifications) against actions (queries, exports, access spikes, sensitive-object touches) and then enforce controls when the story doesn’t hold.

“AI for cybersecurity” starts with a pragmatic framework for making security safer, faster, and more operationally truthful. I work to design cybersecurity systems that hold up under modern deception, financial, operational, or insider-driven, 

My work translates past failures into real controls teams can implement now.