Put the direction in as one prompt
Apply returns the results of Set, Log, and Analyze to a live AI. A group that has just set its hierarchy gets its setting and logging instruction bundled into one system prompt, ready to use now.
AIO 20004: return Set, Log, and Analyze results into the running system
The standard number is the goal statement. The digits 2·00·0·4 say whose values, in which domain, and which action this document covers.
Organization: teams, institutions, communities
Common: domain-independent
Base document
Apply: return the results into the system
What this standard defines: the procedure for combining a setting (20001) and the logging instruction (20002) into one system prompt, the block structure and order of that prompt, and the criteria for choosing a logging output format (inline, structured, tool call). What it does not define: what answers an organization's AI should give. Apply is the procedure that carries a direction; it does not decide the direction (non-normative). It also does not prescribe a model, vendor, or deployment method.
Step 4 of the integrity cycle: Apply
AIO 20001 · Set declares the direction, AIO 20002 — Reasoning Record Standard keeps the record, and AIO 20003 · Analyze finds the drift, and Apply puts the result back into the system. This closes the cycle, which then returns straight to step 1. Apply is not the end point but the start of the next turn.
Agreements die in documents
Most AI governance stops at the document. A team spends half a day agreeing on a hierarchy, produces a well-written principles document, and not one line of it ever reaches the system that actually produces answers. Operations keep running on vendor defaults.
Apply is a practical step that closes that gap. Instead of coding thousands of rules, it turns one agreed direction and one logging instruction into text that can be pasted in. The aim is for even a small organization to complete its first turn of the cycle today.
Because the logging instruction goes in with it, Apply carries its own check. Logs start to accumulate the moment the prompt is in place, and those logs feed the next analysis.
Six blocks of the combined system prompt
The combined prompt is a document with a fixed order. The first five blocks come from the organization's setting; the last comes verbatim from the standard repository.
Identity and scope
The organization and domain (medical, education, public sector, and so on) this setting applies to. The domain maps directly onto the context code in the log.
Value hierarchy
The priority order agreed in the workshop. Not what is right, but what comes first when two goods collide.
Red lines
Items that are never reversed. If the hierarchy is a flexible preference, a red line is a boundary that does not negotiate.
Situational exceptions
Situation red lines and value weights set for the domain. This is what shifts by situation, even inside one organization.
Policy under uncertainty
Whether to hold, escalate to a human, or answer conservatively when the call is genuinely close. Most incidents come from leaving this blank.
Logging instruction (AIO 20002)
The logging prompt, verbatim from the standard repository. It has the model leave one C/V/E/S line per decision, which the end user never sees.
The sixth block is used exactly as committed to the standard repository AIO20002. The site never rewrites it, and the commit hash records which version went in.
Three logging output formats
Apply involves only one real choice: how the model emits the log. All three formats carry the same C/V/E/S code, and none of them shows the log to the end user.
Inline tag
Adds one <aio20002_log> line at the end of the response. It works in any chat interface, so it is the easiest place to start.
Trying it today
Structured output
Splits the response and the log into JSON fields. Best when a system has to parse and store the log.
Wiring into a product
Tool call
The model records the log through a record_aio20002_log tool call. The log stays fully separate from the response, so the risk of leakage is lowest.
Standing up an audit trail
What is available now, and what comes once logs accumulate
Setting + logging = one system prompt
‘Start’ at the top of the page walks through (1) picking the setting to apply, (2) choosing the logging response format, and (3) downloading the combined system prompt as .md or copying it to the clipboard. After sign-in, saved 20001 projects appear in the list.
Without a setting yet, the AIO 20001 workshop produces one in 30 minutes.
Settings-versus-log check, risk check, improved prompt
Once logs accumulate, AIO checks whether the setting matches the logged results and runs an additional benchmark (20003) to look for risk signals. Those findings go into a finer setting, returned as one improved system prompt.
What this step does NOT claim
- Putting the prompt in does not guarantee the model follows the direction. Checking that it does is the job of Log (20002) and Analyze (20003); Apply only means something with those two in place.
- A system prompt is not a safety mechanism. It does not replace guardrails, filters, or human review; it adds a direction on top of them.
- The log the model leaves is a self-report. It is not a causal trace of internal computation, but a record of what the model reports it prioritised.
- The Apply standard document (20004) is not final. What is offered today is a stage-1 implementation of the procedure; the standard will be finalised through the public RFC.
First, find out how much is standing
Five questions, each answered yes / partially / no. Each ‘no’ points to the step of the cycle that is empty. It takes two minutes, and nothing is stored.
Building application cases together
The Apply step becomes a standard only as real organizational cases accumulate. Which blocks were missing in a given domain, and which output format worked in practice, become the content of the 20004 RFC. The experience of organizations that join as application cases feeds directly into improving the combining procedure.