# Prioritising with a Scoring Model — AI Strategy, Ethics and Governance

Source: https://www.geekswithgeeks.com/en/ai-strategy-governance/strat-prioritise

> Rank candidate use cases by value, feasibility, data readiness and risk.

## Make trade-offs explicit

A **weighted scoring model** rates each idea 1 to 5 on a few criteria and combines them with weights that reflect your priorities. It does not give a perfect answer, but it replaces arguments about favourites with a shared, visible method. Typical criteria: **business value**, **feasibility** (technical difficulty), **data readiness**, and **low risk** (so that a high score means "safer"). Revisit the weights if leaders disagree; that conversation is the useful part.

## Scoring three ideas, run

I ran this. Loan auto-approval has the highest value (5) but scores lowest overall because it is risky and its data is less ready; support reply drafts come out on top at 4.3.

```python
W = {"value": 0.4, "feasibility": 0.3, "data_ready": 0.2, "low_risk": 0.1}
cases = {
  "Support reply drafts": {"value": 4, "feasibility": 5, "data_ready": 4, "low_risk": 4},
  "Loan auto-approval":   {"value": 5, "feasibility": 3, "data_ready": 3, "low_risk": 1},
  "Meeting summaries":    {"value": 3, "feasibility": 5, "data_ready": 5, "low_risk": 5},
}
score = lambda s: round(sum(W[k] * s[k] for k in W), 2)
for name, s in sorted(cases.items(), key=lambda kv: -score(kv[1])):
    print(score(s), name)
```

Output:

```
4.3 Support reply drafts
4.2 Meeting summaries
3.6 Loan auto-approval
```

## Use scores to start a discussion

If a surprising idea ranks first, ask whether the scores or the weights are wrong. The model is a thinking aid; the decision still belongs to people who know the business.

**Quiz:** Why include a "low risk" criterion in the scoring model?

- [ ] Because it is required by Python
- [ ] Risk does not matter
- [ ] To lower all scores equally
- [x] So high-value but dangerous ideas do not automatically win

*Answer:* So high-value but dangerous ideas do not automatically win. Scoring risk alongside value makes safety part of the ranking instead of an afterthought.