# NIST AI RMF: MAP 3.5

> GAMMA-Ray v0.0.155 (beta, for review; every rating is experimental). One line item of AI Risk Management Framework 1.0, subcategories (nist.MAP-3.5): its text, its place in the source, the indicators that cite it and the evidence that reaches it.
> Page: https://gamma-ray.sgit.ai/provisions/nist.MAP-3.5/ · Source: data/provisions.json


Provision nist.MAP-3.5 · AI Risk Management Framework 1.0, subcategories · subcategory

> Processes for human oversight are defined, assessed, and documented in accordance with organizational policies from GOVERN function.

The text is as taken from the source. **In the source:** [https://airc.nist.gov/airmf-resources/playbook/map/#Map%203.5](https://airc.nist.gov/airmf-resources/playbook/map/#Map%203.5) (the subcategory, in the NIST AI RMF Playbook (anchor found in the page); linked 2026-11-10). The instrument: [AI Risk Management Framework 1.0, subcategories](../../instruments/index.md), a United States government work; freely reusable; read 2026-11-10.

## Who cites it

| Indicator | The fact it asks for | Kind of link | The assertion |
|---|---|---|---|
| [`I38`](../../questions/index.md#I38) | High-impact actions need a person's approval | part: partly evidenced | Asserted in the mapping by gamma-ray@web-summit.dinis; the written basis is not recorded yet. |

A link is our assertion, not the instrument's. "Same" carries the evidence to the provision in full. "Part" marks the provision as partly evidenced: the fact is one piece of what it asks. "Negation" points the other way.

## The evidence that reaches it

| Card | Through | This provision is | The evidence | How far to trust it |
|---|---|---|---|---|
| [RiskMandate Publisher](../../spectrum/cards/riskmandate-publisher/index.md) | `I38` | partly evidenced | [Source](https://riskmandate.ai/team/publisher.html) | π Moderate · S1·R5·A1 ×2 |
| [The PrompTugal.news agent team](../../spectrum/cards/promptugal-newsroom-team/index.md) | `I38` | partly evidenced | [Source](https://promptugal.news/CLAUDE.md) | π Moderate · S1·R5·A1 ×2 |
| [AI Grant Consultant](../../spectrum/cards/granter-ai-grant-consultant/index.md) | `I38` | partly evidenced | "The final submission is done by you on the relevant platform" [Source](https://granter.ai/blog/ai-grant-consultant) | π Moderate · S1·R5·A1 ×3 |
| [AISLE platform](../../spectrum/cards/aisle-platform/index.md) | `I38` | partly evidenced | "it opens the pull request and waits for human approval" [Source](https://aisle.com/platform/autonomous-remediation) | π Moderate · S1·R5·A1 ×3 |
| [ByTheLaw](../../spectrum/cards/bythelaw-legal-chat/index.md) | `I38` | partly evidenced | "após ação prévia e expressa do Utilizador na Plataforma" [Source](https://app.bythelaw.com/pt/termos-citius) | π Moderate · S1·R5·A1 ×3 |
| [Claude Code, on your own machine](../../spectrum/cards/claude-code-local/index.md) | `I38` | partly evidenced | [Source](https://code.claude.com/docs/en/permissions) | π Moderate · S1·R5·A1 ×3 |
| [GetVocal AI agents](../../spectrum/cards/getvocal-agents/index.md) | `I38` | partly evidenced | "with human-in-the-loop approval on high-stakes actions" [Source](https://www.getvocal.ai/llms.txt) | π Moderate · S1·R5·A1 ×3 |
| [GitHub Copilot cloud agent](../../spectrum/cards/github-copilot-cloud-agent/index.md) | `I38` | partly evidenced | "require approval from a user with write access before they will run" [Source](https://docs.github.com/en/copilot/responsible-use/copilot-coding-agent) | π Moderate · S1·R5·A1 ×3 |
| [Junie](../../spectrum/cards/junie/index.md) | `I38` | partly evidenced | "Junie CLI will ask for approval from the user" [Source](https://junie.jetbrains.com/docs/junie-cli.html) | π Moderate · S1·R5·A1 ×3 |
| [Manus](../../spectrum/cards/manus-agent/index.md) | `I38` | partly evidenced | "Send, save as draft, or cancel" [Source](https://open.manus.ai/docs/v2/task-lifecycle/confirmation-events.md) | π Moderate · S1·R5·A1 ×3 |
| [Noxus AI Workers](../../spectrum/cards/noxus-ai-workers/index.md) | `I38` | partly evidenced | "Get approval for critical actions" [Source](https://docs.noxus.ai/platform/agents/tools) | π Moderate · S1·R5·A1 ×3 |
| [OutSystems Agent Workbench](../../spectrum/cards/outsystems-agent-workbench/index.md) | `I38` | partly evidenced | "alerts a human user to provide verification or a decision before proceeding" [Source](https://extranetprdintegrations.outsystems.net/Documentation_API/rest/getContentMarkdown/doc?Guid=5332e43d-bd1c-4403-8f46-5cbdaf198318) | π Moderate · S1·R5·A1 ×3 |
| [Replit Agent](../../spectrum/cards/replit-agent/index.md) | `I38` | partly evidenced | "Review the task’s work log, test results, and preview before applying changes." [Source](https://docs.replit.com/features/agent/task-lifecycle) | π Moderate · S1·R5·A1 ×3 |

Touched never means complies.

## Answer it directly

A company can answer this line item itself, in its own file at `/.well-known/fact-ray.json`, with the node `gamma-ray:provision/nist.MAP-3.5` ([how](../../fact-ray/well-known/index.md)). A company that already assesses itself against NIST AI RMF can reuse that work.

[All provisions](../../provisions/index.md)