The question is not whether taxpayers should comply only if SARS is likely to detect a default. They should comply in any event. The narrower practical point is that, for taxpayers with unresolved defaults, the old assumption that SARS was unlikely to find the issue is becoming harder to sustain.
There was a time when many taxpayers could still persuade themselves that an undisclosed tax problem might stay buried indefinitely. That was never good law. But it was plainly part of the practical logic behind many decisions to do nothing.
That practical logic now deserves reconsideration. Not because SARS has become omniscient. It has not. But because SARS’s own published data points to a revenue authority that is more data-driven, more automated, and more effective at turning compliance interventions into real fiscal results than the older stereotype allows.
SARS’s own numbers point in one direction
SARS’s published annual reports show net revenue collections of R1.6867 trillion in 2022/23, R1.7409 trillion in 2023/24 and R1.8553 trillion in 2024/25, with the preliminary 2025/26 outcome crossing the R2.010 trillion mark. Over the same period, compliance-programme revenue moved from R231.8 billion to R260.5 billion and then to R304.0 billion. Prevented impermissible or fraudulent refund outflows moved from R76.3 billion to R112.6 billion and then to R147.9 billion.
The same pattern appears in SARS’s use of data and automation. In 2022/23, SARS reported that 94.70% of standard taxpayer returns were auto-assessed. It later reported that 4 765 753 taxpayers were auto-assessed for the 2024 tax year, up 24.94% from the prior year, using what SARS itself described as vast data sources. SARS has also described concrete data inflows: in October 2024 it said it was already receiving information directly from local crypto exchanges and engaging the FSCA on registered crypto-asset service providers, while its third-party-data framework requires banks and other financial institutions to submit financial and personal information about taxpayers directly to SARS. That matters because it points away from an older model in which SARS depended primarily on what a taxpayer chose to volunteer, and toward a system in which third-party data and automated matching do more of the early work.[1]
None of that proves that every irregularity will be found, or that every revenue gain is the product of enforcement alone. Economic conditions still matter. But on SARS’s own figures, the broad direction of travel is difficult to miss: better compliance yield, stronger refund-risk controls, and a more mature use of data to detect and address non-compliance.
Why that matters for VDP
Tax compliance should not depend on whether a taxpayer thinks SARS will detect a default. That is the principle. But where a taxpayer is already sitting with historic non-compliance, the practical question becomes how that position should now be dealt with. In that context, assumptions about detection are relevant, not as a moral justification, but as part of a sober assessment of risk.
On that sober assessment, doing nothing is becoming harder to defend. Inaction preserves uncertainty. It also leaves timing, scope and narrative in SARS’s hands if the matter is later identified. VDP, by contrast, is a statutory mechanism for dealing with a known problem on structured terms. It is not always the answer in every case. But for many taxpayers with real defaults, the range of rational options is narrower than it once appeared.
The existence of other non-compliance does not solve your own
It is understandable that some taxpayers feel enforcement is not perfectly even, or that more obvious non-compliance may exist elsewhere. That may be true. But it does not resolve the taxpayer’s own position. A motorist stopped for speeding does not avoid the ticket by pointing out that another driver was travelling faster. The existence of other defaults does not neutralise one’s own.
VDP is not a casual form submission
SARS’s published VDP guidance makes the threshold plain. A valid disclosure must be voluntary, must be made in the prescribed form and manner, and must be full and complete in all material respects. Because VDP relief is a statutory concession, the requirements are applied strictly.[2]
That is where the idea of a do-it-yourself VDP often begins to look optimistic. The difficulty is not merely admitting that something went wrong. The difficulty is identifying, correctly and comprehensively, what went wrong, over which periods, under which tax types, in what amounts, and with what legal consequences. SARS’s published guidance places the burden on the applicant to present the default, the affected periods and the supporting detail in a form capable of evaluation.[3]
That is why experienced input matters. A taxpayer acting alone must first perform a legal and factual reconstruction before SARS even begins evaluating the application. The risk is not only that the exposure is understated. The risk is also that the legal character of the default is misunderstood, a related tax type is missed, the period is framed too narrowly, or the disclosure is presented in a way that later proves incomplete. In VDP, the real difficulty is often not deciding to disclose. It is ensuring that the disclosure is legally accurate, factually complete and quantitatively defensible.
Initial acceptance is not permanent comfort
Some taxpayers take comfort from the fact that a VDP application may appear to move through the initial process largely on the papers presented. Even if that is so in a given case, it should not be mistaken for permanent safety. Initial acceptance is not the same thing as a permanent endorsement of the disclosure.
Section 231 of the Tax Administration Act allows SARS to withdraw VDP relief if it is later established that the applicant failed to disclose a matter that was material to a valid disclosure. In that event, payments already made may simply be treated as part-payment of the broader tax debt. That means the quality of the initial application matters not only at the point of acceptance, but also because the disclosure may have to withstand later scrutiny.[4]
Broader direction of travel
The legislative picture is consistent with the same broader trend. The Tax Administration Laws Amendment Act 4 of 2026 inserted a new Chapter XB into the Customs and Excise Act to create a customs-and-excise voluntary disclosure framework. Although the new rule-making power in section 77ZH(1) comes into effect only on a date to be fixed by notice in the Gazette, the enactment itself remains a clear signal that SARS’s compliance architecture is continuing to formalise and expand.[5]
The same broader pattern can be seen in SARS’s recent trust-compliance measures. As discussed in our article on the expansion of the section 210 penalty regime to trusts, the significance often lies less in dramatic new weapons than in the deliberate extension of existing compliance machinery to new taxpayer segments. The point is not that every development is revolutionary. It is that the cumulative direction of travel is becoming difficult to ignore.[6]
Conclusion
The point is not panic. It is that the practical logic behind leaving historic non-compliance alone has weakened. On SARS’s own published data, the institution looks more capable, more data-enabled and more effective than the older stereotype suggests.
For taxpayers with real defaults, the question is increasingly not whether inaction feels comfortable, but whether it remains rational. Where VDP is the sensible route, it should be approached for what it is: a technical risk-management exercise that needs to be scoped, quantified and presented properly the first time.
[1] SARS Annual Reports 2022/23, 2023/24 and 2024/25; SARS, “SARS crosses historic R2 Trillion mark in Net Revenue Collection in our democracy” (1 April 2026); SARS, “Media release: Preliminary figures on 2024/2025 Filing Season” (31 January 2025); SARS, “Media release: SARS Warns About Crypto Asset Compliance” (9 October 2024); and SARS, “Third-Party Data”.
[2]SARS, “Voluntary Disclosure Programme (VDP)”; South African Revenue Service, Guide to the Voluntary Disclosure Programme.
[3]South African Revenue Service, Guide to the Voluntary Disclosure Programme; SARS, Voluntary Disclosure Programme – External Guide.
[4]Tax Administration Act 28 of 2011, s 231; SARS, “Guide to Voluntary Disclosure Programme”.
[5]Tax Administration Laws Amendment Act 4 of 2026, especially the insertion of Chapter XB into the Customs and Excise Act 91 of 1964 and the commencement provision relating to s 77ZH(1).
[6]SARS, “Implementation Date of Administrative Non-Compliance Penalties for Late- or Non-Submission of ITR12T Income Tax Returns for Trusts” (7 April 2026); SARS, “Final demand issued for annual income tax returns for Trusts” (9 February 2026).
