You can afford to ask
more than one question.
Start with one proposed consequence. Establish the baseline once. Then change the evidence, authority, connection or reality and run another bounded route without buying the whole engagement again.
Keep your architecture. Test the boundary.
Registration and the standing institutional baseline remain distinct from paid professional examination. Explore bounded demonstrations, interoperability, evidence review and execution-readiness work.
EXPLORE AI GOVERNANCE ENGAGEMENTS →What are you trying to make happen safely?
Say it in plain language. Everything else is built around the consequence you want examined.
One consequence. Multiple realities.
Allow an AI agent to change an HVAC setpoint can be examined under more than one condition. Change one bounded fact and watch where the demonstration route stops.
You have $2,500. What can you examine?
Pick a published assessment price and see what that same amount of money could fund under the current TA-14 planning model. The services are different; this is a budget comparison, not an equivalence claim.
Published external budget
Cloud Sentry publishes this engagement at $2,500. Its stated scope is summarized below from the provider's own page.
BUDGET
What that budget can fund here
This is a budget illustration, not a claim that one provider engagement equals 48 TA-14 routes. Methods and deliverables differ.
Published one-time assessment of governance capability, decision-making, data and access.
VIEW PUBLISHED SOURCE ↗Published 1–2 week discovery, data-flow, credential and logging review with findings and remediation roadmap.
VIEW PUBLISHED SOURCE ↗Published standard rate for decision inventory, risk-tier map, approval paths, evidence design and roadmap.
VIEW PUBLISHED SOURCE ↗Change the condition. Run it again.
The point is not to manufacture more work. It is to make it economically reasonable to examine the routes that could materially change the answer.