AI Master Class: SARS Verifications - Automating Workflows
Duration: 2 hours
Price: R195.00
AI Master Class: SARS Verifications - Automating Workflows
Presenters : Niël Malan
Overview
Every week, tax practices lose billable hours to the same work: responding to verification requests, corresponding with SARS, handling objections and requests for reasons, and sorting client records that arrive in no particular order.
Most practitioners already have generative AI on their desks. The harder part is knowing how to brief it well enough to produce work worth reviewing, and then checking that work without giving back the time saved.
Rather than covering the method in theory, this video works through a complete case study. It begins with a fictitious company, its supporting records and a SARS verification letter received through eFiling, and ends with a reviewed response that is ready for submission. Practitioners receive the AI skill used in the demonstration, which was built for South African tax practice, together with a brief look under the hood. The skill turns a loose request into a properly briefed one without requiring the practitioner to manage the mechanics.
The demonstration runs twice. In the junior version, the skill explains what a verification request means, why each document belongs in the response pack and what the reviewing senior checks. It drafts the response while explaining the work, so that trainees continue to develop their judgement. In the senior version, the same skill produces the work product, deadline table and review checklist. The checklist looks for four ways in which AI gets South African tax law wrong: invented authority, positions superseded by amendments, law imported from other jurisdictions, and unsettled issues presented as settled. The practitioner reviews and signs the final response.
The session comprises 90 minutes of content and 30 minutes of open discussion. Niel Malan, an engineer, explains what an AI skill is, how these systems succeed and fail, and how their output is verified.
The session is directed at practices as well as individual practitioners. It does not recommend a product or endorse a vendor. The approach applies to whichever generative AI platform a practice already uses, and professional judgement remains with the practitioner throughout.
Video Content
- Part 1: Where the hours go — the work a compliance practice hands to AI, and the work it always keeps with the practitioner.
- Part 2: What a skill is — a brief look at good instruction, and why it is built into the skill instead of taught as a separate method.
- Part 3: The case study, junior seat — a SARS verification letter explained, supported with evidence and drafted, while the skill coaches rather than replaces the trainee.
- Part 4: The case study, senior seat — the faster route from letter to reviewed response pack, with deadlines calculated and diarised.
- Part 5: Trust, then verify — the four ways in which AI gets South African tax law wrong, and the review that makes its output safe to rely on.
- Part 6: Practice guardrails — client confidentiality, limits on authority, and what stays with the practitioner.
- Q&A and open discussion (30 minutes)
Competencies Developed
On completion of the video, practitioners are able to:
- Identify AI-suitable compliance work and distinguish tasks that are delegated to AI from those requiring professional judgement.
- Structure effective AI instructions that produce relevant, practice-ready outputs from unstructured client information.
- Interpret SARS verification requests and identify the supporting documentation and evidence required for a complete response.
- Use AI to draft compliance responses while maintaining appropriate practitioner oversight and review.
- Apply AI in both junior and senior workflows, using it for trainee development, work-product preparation, deadline management and review.
- Verify AI-generated tax content by identifying invented authority, outdated legislation, foreign law and unsettled tax positions presented as established law.
- Evaluate AI-generated work critically using a structured review and verification process before submission to SARS.
- Apply appropriate practice guardrails relating to client confidentiality, data handling, professional responsibility and the limits of AI.
- Assess the impact of AI on practice efficiency, including review capacity, trainee development and the economics of compliance work.
- Integrate AI into existing practice workflows without replacing professional judgement or accountability.