THREE TIERS

One framework. Three tiers.
So you can start where you are.

EW-AiRM™ scales from a single AI system start-up to a global enterprise running hundreds of deployments. The framework is identical throughout; only the depth of its implementation changes.

'Core' is not a lesser EW-AiRM^TM version and 'Full' is not the "real deal" everyone migrates to: The tier is a match to organisational reality, not a maturity ladder.

Core

ENTRY

Start-/scale-ups and limited AI deployment:

<100 employees
1–5 AI system

Six pillars applied at conceptual level

HAiPECR^TM run as a pre-deployment filter

Top-tier MIT risks only (Primary mappings)

One named accountable owner per system

Quarterly review cycle

Self-assessment using the Maturity Tool

Adopt Core →

Standard

RECOMMENDED

For mid-market firms and regulated SMEs:

100–1,000 employees
5–25 systems

All six pillars with documented evidence

HAIPECR^TM pre-deployment + continuous review

Primary + Secondary MIT mappings

Two-line accountability (system + risk owner)

Six AI Black Swan categories rehearsed

Monthly KXI dashboard, quarterly board pack

Adopt Standard →

Full

ENTERPRISE

For global enterprises and systemically important institutions:

1,000+ employees
25+ systems)

All six pillars with second-line review

HAiPECR^TM audit trail per deployment

Full Primary / Secondary / Tertiary mappings

Three-line accountability + board sponsor

All eight Black Swan categories annually

Weekly KXI, monthly risk, quarterly board

Adopt Full →

FLOOR CONDITIONS: The 5 Non-Negotiables

Five things every tier must do.

Named accountability for every deployed AI system

HAiPEICR^TM run as documented pre-deployment filter

Human override capability tested with a 4-hour SLA

Documented board or executive risk acceptance

Incident reporting pathway, with named recipient

You can use our tiering and the assessment tools as diagnostic (you already have AI deployed), as a roadmap (you are about to deploy), or as a reference guide (you have specific governance problems).

Not sure which tier fits?

The Maturity Assessment Tool will tell you in 20 minutes. Free, runs in your browser, nothing leaves your machine.

Run the assessment →

Talk to us →

人类人工智能学院 是一个独立的思想/行动智库，召集关于人工智能治理的多边和多方利益相关者对话。该研究所与联合国、联合国教科文组织、经合组织、欧盟委员会和世界各国政府进行合作。

之家 EW-AiRM™ — 一个开放的、三层的企业级人工智能风险管理框架，基于公开许可的标准，并与联合国教科文组织、联合国欧洲经济委员会、美国国家标准与技术研究院、国际标准化组织和欧盟人工智能法案保持一致。

咨询、培训与建议

商业咨询、培训和咨询服务由以下机构提供： De-Risking Solutions Ltd. 和 RiskAi.Ai

人工智能与技术解决方案

人工智能、代理工具和技术解决方案由...提供 Human-Ai.Solutions.

组织

人类人工智能学院

由...运行和维护
De-Risking Solutions Ltd.

电子邮件：
contact [at] human-ai.institute

关注我们：

人工智能学院

专题页面

YouTube 频道

已列入经合组织人工智能政策观察站 · 联合国大学人工智能网络创始成员 · 符合教科文组织、UNECE WP.6、NIST 和 ISO 42001 标准

One framework. Three tiers. So you can start where you are.

'Core' is not a lesser EW-AiRMTM version and 'Full' is not the "real deal" everyone migrates to: The tier is a match to organisational reality, not a maturity ladder.