Zero Trust for a 50-Person Firm: What It Means and What It Costs

Zero trust isn’t a product. For a 50-person firm it is five decisions about what the network stops vouching for, mostly made with licenses already paid for. Here is what it means, in NIST’s terms, and what it costs.
How to Evaluate an MSP’s Security Claims: The Evidence List to Ask for Before You Sign

Two MSP proposals, the same promises, and no evidence in either. The ten documents, settings and contract terms I’d ask any managed service provider for before handing over the keys to the business.
AI Agents Are Nonhuman Users: Identification and Authentication Under NIST SP 800-171

An AI agent that reads your mailbox or posts to your ERP is a process acting on behalf of a user, and NIST SP 800-171 already covers it. Six rules for giving agents their own identity, credentials, scope and log.
The Three AI Questions a Board Should Ask Every Quarter, and the One-Page Answer

A board doesn’t need a longer AI slide. It needs three questions every quarter and a one-page answer with the inventory, data map, controls, incidents, decisions and calendar behind it.
The CMMC Executive Affirmation: What You Personally Attest to When Your Company Self-Assesses

When your company self-assesses, a senior official affirms in SPRS that every CMMC requirement is implemented and will stay that way. Here is what that statement says and what to see before you sign it.
What Is Shadow AI? A CEO’s Definition and the First Inventory You Can Run This Week

Shadow AI is any AI tool, feature or agent that touches company or client data without a company decision behind it. Here is the definition I give CEOs and a five-day inventory you can run with tools you already own.
The SAFE Framework: Four Steps to Govern AI Without a Chief AI Officer

Surface, Aim, Fence, Expand: the four-step sequence I teach leadership teams for getting control of the AI already in use in their companies, without hiring a chief AI officer.