1---
 2layout: page
 3title: Formalize Sensitive Activities as a Layer
 4---
 5
 6- **ADR:** 0009
 7- **Proposal Author(s):** @eddie-knight
 8- **Status:** Accepted
 9
10## Context
11
12We have often recieved feedback regarding the lack of a "Layer" to describe sensitive activities that live between Policy and Evaluation.
13
14- Layer 1: Guidance
15- Layer 2: Controls
16- Layer 3: Policy
17- (No Layer) Sensitive Activities
18- Layer 4: Evaluation
19- Layer 5: Enforcement
20- Layer 6: Audit
21
22The argument has been that since we don't provide schemas for the diverse array of possible sensitive activities, we shouldn't assign a layer to it. However, as we move closer to a versioned release of the model, it has become clear that the model is not dependent on having schemas for every component.
23
24## Decision
25
26We will assign a layer to sensitive activities in the model.
27
28## Consequences
29
30- This will require +1 renumbering of layers 4, 5, and 6 (to 5, 6, and 7 respectve)
31  - Layer 4: Sensitive Activities
32  - Layer 5: Evaluation
33  - Layer 6: Enforcement
34  - Layer 7: Audit
35- No impact on schemas and SDKs, as they have already removed any reference to numberic layer identifiers
36- All documentation and web content must be updated
37- Adopters of previous versions will be disrupted by this change in terminology
38
39## Alternatives Considered
40
41We have previously ruled this out due to the intended stability of the model. However, with the upcoming release of the whitepaper, change of repository locations, and major (v1) changes to the schemas and SDKs, this is an opportunity for a breaking change to the model which will be included in the official whitepaper for long term stability.