Kevin Harbauer
Transformation & Growth Executive
I have spent more than 20 years leading product, engineering, and security teams in environments where delivering results was not optional. CTO, CISO, and board advisor across PE-backed software, healthcare, and legal services. I write here about the work that is actually hard — technology leadership, AI adoption, security, and building organizations that execute.
Recent Articles
AI Watching AI
At scale, humans can't review every agent interaction. The case for guardian agents — and why AI overseeing AI is uncomfortable but probably inevitable.
Time to Hello World
Vibe coding made getting to a first API call table stakes. The harder question: was your documentation written for the agent reading it for a user?
Built Secure, Deployed Dangerous
An agent built correctly can still drift into dangerous territory through misconfiguration. Most organizations have no way to detect it until something breaks.
Not Every Decision Needs a Human... But Some Do
Most AI agents run at whatever autonomy level was easiest to implement, not the one that reflects actual risk. Here's how to tell the difference.
The Agent Can Think. It Just Can't Get to the Data.
The biggest barrier to real AI automation isn't the model. It's connectivity. And the protocol solving it is creating your next governance problem.
Your Digital Workforce Has a Landlord
Thousands of AI agent workflows disrupted overnight — not because agents broke, but because a vendor changed its billing. That's a governance failure.