---
id: "framework-standard-rai-model"
type: "framework"
source_timestamps: ["§ Three Flaws of the Standard Approach", "¶1"]
source_url: "https://hbr.org/2026/05/what-are-your-companys-ai-nightmares"
source_title: "What Are Your Company's AI Nightmares?"
tags: ["legacy-models", "governance"]
related: ["concept-standard-rai-approach"]
steps: ["\\\"Articulate organizational AI ethics values (e.g.", "fairness", "privacy", "transparency", "accountability", "safety).\\\"", "\\\"Translate those abstract values into enterprise-wide procedures (e.g.", "checking for bias", "filtering sensitive data).\\\"", "\\\"Enshrine those procedures into a formal", "enterprise-wide AI policy.\\\"", "Implement the policy across the organization.", "Create a Responsible AI board (or assign to an existing risk board) to handle escalations of high-risk AI cases."]
sources: ["governance"]
sourceVaultSlug: "hbr-seg-governance"
originDay: 7
articleStem: "hbr-cl-82-ai-nightmares"
sourceUrl: "https://hbr.org/2026/05/what-are-your-companys-ai-nightmares"
sourceTitle: "What Are Your Company’s AI Nightmares?"
---
# The Standard Responsible AI Implementation Model

The traditional, **sequential** process enterprises use to establish AI governance. Blackman ([[entity-reid-blackman]]) presents it specifically to *critique* it — for being slow, vague, and hard to communicate.

**The five steps:**
1. **Articulate values** — organizational AI ethics values (fairness, privacy, transparency, accountability, safety).
2. **Translate to procedures** — turn abstract values into enterprise-wide procedures (checking for bias, filtering sensitive data).
3. **Enshrine in policy** — codify the procedures into a formal, enterprise-wide AI policy.
4. **Implement** — roll the policy out across the organization.
5. **Create a board** — stand up a Responsible AI board (or reuse an existing risk board) to handle high-risk escalations.

Each step maps to a flaw: Step 3-4 → too slow ([[claim-standard-rai-too-slow]]); Step 1-2 → no clear success metric ([[claim-values-wrong-start]]); Step 3 → the [[quote-tower-of-babel]] communication failure. This model is the conceptual foil for [[framework-enc-questions]]; the full narrative lives in [[concept-standard-rai-approach]].

**Enrichment note:** This sequence matches the mainstream principle-first Responsible AI pattern used by major technology firms, standards bodies, and regulators — Blackman is not misrepresenting it, he is contesting its *ordering* (values-first vs. nightmares-first).


## Related across articles
- [[framework-ai-risk-oversight]]
- [[framework-board-cyber-engagement]]