Table 3 Summary of potential research... | Emerald Publishing

Table 3

Summary of potential research questions

Theme	Challenged accounting Construct(s)	Theory-driven future research questions
Audit evidence, materiality and inference in an algorithmic environment	Audit evidence; materiality; going-concern assessment	What constitutes “audit evidence” when key inputs are probabilistic, model-generated signals (e.g. sentiment scores, anomaly detections, satellite imagery) that are not directly observable or reproducible by human auditors? How do algorithmic forms of inference challenge traditional epistemic assumptions underpinning evidence sufficiency, corroboration and audit trails? How should materiality and going-concern assessments be reconceptualised when AI systems identify statistically significant risks that may fall outside established professional or regulatory relevance thresholds?
Professional judgement, scepticism and accountability under hybrid intelligence	Professional judgement; professional scepticism; accountability	In AI-mediated professional judgement, who is responsible for identifying and challenging error, and what mechanisms ensure that algorithmic outputs are subject to appropriate human scepticism and oversight? What does “professional scepticism” mean when judgement is partially delegated to opaque or probabilistic AI models, and how can scepticism be exercised towards AI agents? What forms of expertise distinguish professional judgement from technical model operation in hybrid intelligence accounting, and how does this distinction underpin the profession's claim to epistemic authority?
Reasonable assurance, jurisdiction and legitimacy in AI-enabled accounting	Reasonable assurance; professional jurisdiction; legitimacy	How should the concept of “reasonable assurance” be redefined when audit conclusions depend on continuously learning, non-deterministic AI systems rather than stable procedures and documentation? Can existing auditing and reporting standards accommodate AI-based inference, or do algorithmic systems expose foundational limits in standards built around human judgement and explainability? Who should hold epistemic and legal authority over AI-mediated accounting judgements (such as professional bodies, regulators, or technology vendors) and what are the implications for the future jurisdiction of the accounting profession?

Theme	Challenged accounting Construct(s)	Theory-driven future research questions
Audit evidence, materiality and inference in an algorithmic environment	Audit evidence; materiality; going-concern assessment	What constitutes “audit evidence” when key inputs are probabilistic, model-generated signals (e.g. sentiment scores, anomaly detections, satellite imagery) that are not directly observable or reproducible by human auditors? How do algorithmic forms of inference challenge traditional epistemic assumptions underpinning evidence sufficiency, corroboration and audit trails? How should materiality and going-concern assessments be reconceptualised when AI systems identify statistically significant risks that may fall outside established professional or regulatory relevance thresholds?
Professional judgement, scepticism and accountability under hybrid intelligence	Professional judgement; professional scepticism; accountability	In AI-mediated professional judgement, who is responsible for identifying and challenging error, and what mechanisms ensure that algorithmic outputs are subject to appropriate human scepticism and oversight? What does “professional scepticism” mean when judgement is partially delegated to opaque or probabilistic AI models, and how can scepticism be exercised towards AI agents? What forms of expertise distinguish professional judgement from technical model operation in hybrid intelligence accounting, and how does this distinction underpin the profession's claim to epistemic authority?
Reasonable assurance, jurisdiction and legitimacy in AI-enabled accounting	Reasonable assurance; professional jurisdiction; legitimacy	How should the concept of “reasonable assurance” be redefined when audit conclusions depend on continuously learning, non-deterministic AI systems rather than stable procedures and documentation? Can existing auditing and reporting standards accommodate AI-based inference, or do algorithmic systems expose foundational limits in standards built around human judgement and explainability? Who should hold epistemic and legal authority over AI-mediated accounting judgements (such as professional bodies, regulators, or technology vendors) and what are the implications for the future jurisdiction of the accounting profession?