As the SARS collection process becomes more technologically advanced, tax practitioners must advise clients to avoid tax evasion and remain compliant.
This is according to Mahomed Kamdar, Tax Specialist at the South African Institute of Professional Accountants (SAIPA). “SARS will soon become so efficient at detecting tax evasion and recovering hidden revenues, it will no longer be worth it to lie on returns,” he says.
For example, to qualify as a small business corporation and obtain tax relief from SARS, a company’s shareholders must hold no shares in any other company, locally or internationally. If they do but apply for relief anyway without disclosing this fact, this information can be easily discovered electronically.
Similarly, a company is disqualified from seeking tax relief from SARS if a discretionary trust owns shares in it. Even if the taxpayer fails to reveal such shareholding on application, SARS’ IT systems can easily uncover it through high-speed cross-referencing.
In his budget speech, Finance Minister Mboweni allocated R3 billion to SARS to ensure its collection efforts succeed.
SARS may be utilising artificial intelligence (AI) to identify false submissions. But how does it work?
AI is a form of computer programming. In traditional programming, a computer receives data (inputs). These are processed by steps predefined by a programmer to produce some desired result (outputs).
In the case of AI, a data scientist gives the computer a vast collection of known inputs and outputs. Its AI software crunches this data to figure out how those inputs resulted in the given outputs. It effectively programmes itself, producing a reusable procedure or “model”. That model can be applied to new inputs to identify those that would produce the same outputs.
In the case of tax, an AI system can develop a model for recognising false submissions based on its analysis of ones belonging to known tax offenders. In the era of smart machines, the pattern of tax evasion becomes a fingerprint by which deceptive returns can be identified with high probability.
As SARS gains access to third-party systems, like property and vehicle registries, the chance of being caught out increases exponentially. “You don’t need AI to notice that if a taxpayer pleads poverty on their return but bought a luxury vehicle or house that year, something doesn’t add up,” notes Kamdar.
SARS is not only upgrading its internal processes. It is now capable of seamlessly sharing data with authorities around the world to discover foreign caches of undeclared income.
Overall, global tax collection is becoming smarter and more collaborative, leaving dishonest taxpayers with fewer options for hiding either hard-earned or ill-gotten gains.
This is highlighted by the fact that SARS recently advertised about ten vacancies, all of them for positions specialising in IT. In fact, there is speculation throughout global companies that many top IT professionals are now in the employ of tax authorities around the world.
In addition, the Income Tax Act now empowers SARS to act not just against intentional evasion but also negligence on the part of the taxpayer.
Tax practitioners have a duty to advise their clients on these trends and convince them to be completely transparent in their declarations and deductions. This means they too must become knowledgeable about what SARS is doing and educate themselves on how technology can disrupt the status quo.
SAIPA’s continuous professional development (CPD) programmes now focus heavily on emerging technologies, like AI and data analytics, for all its tax designations. This includes Professional Tax Specialist (SA), Professional Tax Practitioner (SA) and Professional Tax Technician (SA).
“The Institute’s goal is to prepare all its members for the challenges and opportunities of the Fourth Industrial Revolution as future-ready accountancy professionals,” says Kamdar.
SARS’ technology drive is a good example of how 4IR will affect trade, industry, society and governance.