BMAN30030 Chap.10 Generative AI and Professional Judgement
Generative AI and Professional Judgement
Generative AI and Professional Judgement asks what happens to the expertise that justifies the accounting profession when generative tools take over routine work. The readings combine different kinds of evidence. Lee et al.
(2025) surveyed 319 knowledge workers and found that confidence in the tool is associated with less critical thinking, while self-confidence is associated with more; they also show effort moving from gathering information to verifying it, and from doing tasks to overseeing them. Samiolo et al.
(2023) interviewed auditors, regulators and software providers and argue that apparently simple audit tasks carry judgement that trains early-career staff. Salijeni et al. (2021) show how big data analytics widen audit evidence and help auditors justify conclusions while reshaping relationships inside firms. Consultancy surveys add the scale of adoption and the gap between leaders' perceptions and staff use.
The topic has a practical side for this module, because the final examination permits restricted AI use for research under category C and requires a declaration form, accurate citations and a visible authorial voice. Good answers weigh efficiency against the erosion and relocation of judgement, rather than predicting a simple future.
What this chapter covers
- 01
Generative AI and the meaning of professional judgement
- 02
Confidence, critical thinking and cognitive effort
- 03
Automation of routine audit work and learning on the job
- 04
Big data analytics in financial statement audits
- 05
Survey evidence on adoption and perception gaps
- 06
Category C AI rules in the final examination
Worked example · free
Judge an AI-assisted answer plan against the academic integrity expectations
- 1Locate and read the original sources, because the guidance requires originals of all cited sources to be reviewed and AI tools can fabricate citations.
- 2Rewrite the paragraph so that the argument, evaluation and conclusion are the student's own, preserving authorial voice rather than assembling quotations.
- 1Record the prompts and checks used and declare the use on the compulsory acknowledgement form, citing any AI output relied on.
Key terms
- Professional Judgement
- Professional judgement is the application of trained knowledge and experience to reach a defensible decision where rules alone do not settle the answer.
- Task Stewardship
- Task stewardship is the oversight work of directing, checking and taking responsibility for output that a tool has produced.
- Authorial Voice
- Authorial voice is the visible presence of a writer's own position, interpretation and reasoning in submitted work.
Generative AI and Professional Judgement FAQ
Does generative AI make accountants' judgement less important?
The readings suggest judgement moves rather than disappears. Survey evidence shows effort shifting towards verifying and integrating AI output, and interview studies of audit argue that routine tasks remain the place where junior staff learn to deliberate. Removing that work without redesigning training could weaken the expertise firms rely on later.
Can I use AI in the final examination?
The examination is classed as category C, permitted use, which allows generative AI in a restricted form for research. You must review the originals of every source you cite, cite AI output you rely on, keep a record of your use and submit the declaration and acknowledgement form. The paper can be completed without AI.
Which readings support arguments about audit technology?
Samiolo et al. (2023) use 44 interviews to question the idea that low-level tasks are judgement-free, and Salijeni et al. (2021) use 25 interviews to show how analytics scripts and dashboards change audit evidence and communication. Lee et al. (2025) add survey evidence on critical thinking among knowledge workers more generally.
Exam move
Organise this chapter around evidence types. Make a table with one row per reading and columns for method, sample, main finding and limitation: Lee et al. (2025) for survey evidence on critical thinking, Samiolo et al. (2023) for interview evidence on audit judgement, Salijeni et al. (2021) for analytics in practice, and the McKinsey survey for adoption at scale.
Writing the limitation column first forces you to read critically, which is exactly what the first-class descriptor rewards. Next, practise two opposing paragraphs: one arguing that generative AI strengthens judgement by freeing time and widening evidence, the other arguing that it erodes the learning that produces judgement.
Then write a third paragraph that reconciles them with conditions, such as training in verification or redesigned junior roles. Link the topic to earlier weeks: professional bodies' claims to serve the public interest and the oversight debate both depend on trust in expert judgement.
Because past papers have asked whether the profession needs radical or incremental reform in the face of technology, prepare a thesis on that question too. Finally, treat the module's own AI rules as revision content: know category C, the AMBS expectations of accuracy, authorial voice and transparency, and the declaration form, and keep an audit trail of any AI use while you revise.
In the final fortnight, rehearse one timed paragraph a day in which you argue a position on AI and audit quality using two readings and one current example. Afterwards, check each citation against the original text and note what verification changed, so the exam habit of reviewing sources becomes automatic.
Working through Generative AI and Professional Judgement in BMAN30030? Sia is AskSia’s AI Accounting tutor — ask any BMAN30030 Generative AI and Professional Judgement question and get a clear, step-by-step explanation grounded in how BMAN30030 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.