Table A4

AI system classes mapped to DIKW layers and decision phases, with typical autonomy risks and safeguards

AI classPrimary DIKW contributionDecision phases most affected (frame/evaluate/commit/ enact)Typical autonomy risksAutonomy-preserving safeguards
Predictive/descriptive (ML and data mining)Data → information (aggregation, patterning)Frame, evaluateOver-weighting historical patterns; metric myopia; false objectivityData governance and provenance; counter-metric reviews; alternative framing prompts; sandbox scenario tests
Prescriptive/optimization and recommendersInformation → knowledge (action suggestions)Evaluate, commitOptimization without values; goal mis-specification; automation biasValue/constraint reviews; multi-objective criteria; opt-out and override protocols; red-team “what-ifs”
Generative AI assistants (LLMs)Information ↔ knowledge (summarize, ideate, articulate)Frame, evaluate, enactFluent but unfounded rationales; leakage of internal knowledge; prompt steeringSource citation/RAG; confidentiality controls; traceable prompts; justification checklists
Perception and extraction (NLP/CV/OCR)Data formation (from unstructured inputs)Frame, enactGarbage-in; biased extraction; loss of contextHuman validation on samples; bias audits; retention of raw context; escalation rules
Source(s): Authors’ own work

or Create an Account

Close Modal
Close Modal