Table 2 Conceptual framework for hybrid... | Emerald Publishing

Table 2

Conceptual framework for hybrid intelligence accounting

Phase of the accounting knowledge process	Machine role	Human role	Jurisdictional implications
1. Evidence generation	Detect patterns and anomalies Update continuously Determine relevance and materiality	Ensure data governance, quality and provenance Identify gaps, context loss or model blind spots	Expands the definition of audit and accounting evidence; expertise shifts towards oversight of data provenance and evidentiary boundaries
2. Inference and prediction	Model nonlinear relationships Generate probabilistic forecasts Identify latent structures and risks	Interpret outputs in economic and organisational context Test robustness, reasonableness and stability Integrate domain knowledge and standards	ML functions as a parallel inference system, challenging the profession's traditional primacy over diagnostic reasoning
3. Interpretation and judgement	Propose candidate explanations Highlight salient variables Generate scenario and sensitivity simulations	Apply professional scepticism Weigh heterogeneous evidence Consider ethical, regulatory, and strategic consequences	Re-centres professional authority in interpretive judgement; hybrid reasoning becomes core to maintaining jurisdiction
4. Governance and accountability	Provide audit trails and documentation Monitor drift, bias and model integrity Implement fairness and compliance checks	Set governance standards for AI systems Validate model design, assumptions and inputs Communicate limitations and assign responsibility	Positions accountants as governors of AI systems rather than sole producers of judgement, preserving jurisdiction through oversight and accountability mechanisms

Phase of the accounting knowledge process	Machine role	Human role	Jurisdictional implications
1. Evidence generation	Detect patterns and anomalies Update continuously Determine relevance and materiality	Ensure data governance, quality and provenance Identify gaps, context loss or model blind spots	Expands the definition of audit and accounting evidence; expertise shifts towards oversight of data provenance and evidentiary boundaries
2. Inference and prediction	Model nonlinear relationships Generate probabilistic forecasts Identify latent structures and risks	Interpret outputs in economic and organisational context Test robustness, reasonableness and stability Integrate domain knowledge and standards	ML functions as a parallel inference system, challenging the profession's traditional primacy over diagnostic reasoning
3. Interpretation and judgement	Propose candidate explanations Highlight salient variables Generate scenario and sensitivity simulations	Apply professional scepticism Weigh heterogeneous evidence Consider ethical, regulatory, and strategic consequences	Re-centres professional authority in interpretive judgement; hybrid reasoning becomes core to maintaining jurisdiction
4. Governance and accountability	Provide audit trails and documentation Monitor drift, bias and model integrity Implement fairness and compliance checks	Set governance standards for AI systems Validate model design, assumptions and inputs Communicate limitations and assign responsibility	Positions accountants as governors of AI systems rather than sole producers of judgement, preserving jurisdiction through oversight and accountability mechanisms