AI Investment Checklist

Project Information

Initiative Name

Brief Description

—

Three Horizons Classification

Portfolio targets: H1 60-70%, H2 20-30%, H3 10-15%

AI Transformation Matrix Quadrant

Does this address a strategic gap in our portfolio?

Total Investment (3 years)

Total Expected Benefits (3 years)

Sum of: Cost Reduction + Revenue Growth + Risk Reduction

3-Year ROI

Resources available (budget, people, time)

Executive sponsorship secured

Timeline realistic (not over-optimistic)

Dependencies identified and manageable

—

Score each factor from 0-10. Weighted total must be ≥7 to proceed.

Factor	Score (0-10)	Weight	Weighted
Data Availability Do we have enough historical data?		25%	0.0
Data Quality Is data accurate, complete, consistent?		30%	0.0
Data Access Can we integrate necessary data sources?		20%	0.0
Data Bias Is historical data representative/unbiased?		25%	0.0

Total Data Readiness Score

0.0 / 10

Learning Approach

Static AI
Trained once, doesn't learn from new data (Annual cost: ~10-15% of development)

Batch Retraining
Periodic updates monthly/quarterly (Annual cost: ~25-30% of development)

Real-time Learning
Continuously learns from new data (Annual cost: ~35-50% of development)

Have we budgeted for ongoing retraining costs?

Do we have infrastructure for continuous learning?

Business Impact of AI Error

Expected AI Accuracy (%)

Have we defined acceptable failure rate?

Is human oversight plan defined and budgeted?

Regulatory/Business Requirements

Have we budgeted for explainability tools/processes?

Potential Discrimination Risk

Have we planned for bias testing and auditing?

Have we completed legal/ethics review?