Robust
24+ months
Decision-makers are named, dissent is documented, vendor fallbacks are mapped, and planned 90-day reviews actually run.
Governance shelf-life
Most AI governance asks if you pass today. I ask how long until it breaks.
Your governance has a shelf-life. The free AI Governance Durability Check™ tells you which structural element expires first, and what durability would look like for an organization like yours.
The problem
Governance built on someone else's reference can become wrong overnight.
Most AI governance is referenced — built on regulators' interpretations, vendors' policies, specific models, frameworks like NIST AI RMF or ISO 42001, or an outside counsel's memo. When the reference moves, governance built on it can become wrong overnight without anyone noticing.
Governance decisions read fine on paper. The reference material changes. Now what?
How this is assessed: Methodology.
What 'shelf-life' means
Four categories. Worst cause wins.
The diagnostic estimates which category your governance sits in across eight structural causes. Your overall shelf-life is set by the worst one — not the average. Governance lasts as long as its weakest structural link.
Functional
12–24 months
Structures exist on paper but have not been exercised. The path is documented; nobody has used it yet.
Fragile
6–12 months
Informal channels. No fallback plans. Reasoning lives in two people's heads. One staff change away from a problem.
Critical
under 6 months
No documented escalation. Heavy vendor lock-in without alternatives. Nobody has stress-tested the process.
How it works
A few questions. An honest result, on the spot.
-
Step 1
Answer the questions.
About how decisions get made, how reasoning is preserved, who has authority, and what your governance is anchored to. About ten minutes, in your browser.
-
Step 2
See your durability band.
Brittle, Aging, or Durable-ish — set by your weakest structural cause, not the average. Every part of the result traces back to a specific answer. Nothing is stored unless you ask me to.
-
Step 3
Read what sets the date.
Your three most fragile sections, the first trigger that would void a control, and a 90-day plan. Useful even if you never hire me. When the gaps are structural, the paid Shelf-Life Assessment is the next step.
Engagement options
Three layers. Pricing you can see.
Pricing is segmented, not a range. Tier 2 has five fixed-fee sub-tiers scoped by organization size and sector. Tier 3 is custom advisory or a fractional retainer.
Tier 1
AI Governance Durability Check
Free
Self-serve, instant result
Any organization with AI systems in active use.
Read more →Tier 2
ModalDiagnostic-to-Blueprint
From $750
2-week delivery
Five fixed-fee sub-tiers by org size and sector. Risk map, decision-process RACI, governance framework, implementation roadmap.
Read more →Tier 3
Custom advisory
$12,000+
6 weeks – ongoing
Implementation programs, fractional retainers, public-sector RFPs.
Read more →Who I am
Computational sociologist. Builder of institutional memory.
I'm Chris Graziul — PhD in sociology from the University of Chicago. I've published nineteen peer-reviewed papers and had a position paper accepted at ICML 2025 on AI governance. I work the gap between what frameworks ask organizations to do and what actually happens when models update or the team that wrote the policy leaves.
I also founded the Illinois Data Equity Project — a 501(c)(3) whose own governance is designed to fold into community leadership. It's the methodology applied to its own institution.
Find out how long your governance lasts before something breaks.