Section 11D programme at a glance
Section 11D programme at a glance
By Stephanie Tsumbu, Manager
South Africa’s Section 11D Research and Development (R&D) Tax Incentive offers businesses a valuable opportunity to reduce the after-tax cost of innovation. Where companies undertake qualifying scientific or technological development activities in South Africa, they may qualify for a 150% tax deduction on eligible R&D expenditure. This can include direct people costs, such as salaries and wages for employees involved in R&D, as well as qualifying materials, consumables, equipment usage and other inputs required to carry out the work.
To access the incentive, companies must obtain pre-approval from the Department of Science, Technology and Innovation and include the relevant approval with the annual SARS claim. With the programme now extended until 31 December 2033, businesses have a meaningful window to identify qualifying projects, strengthen their R&D documentation and embed robust processes before expenditure is claimed.
Definition of qualifying R&D
R&D tax incentives are used by many leading economies including Canada, the United States, the United Kingdom, Japan and Portugal. Although each jurisdiction applies its own legislative framework, the core policy objective is broadly consistent: to encourage businesses to invest in activities that advance scientific or technological knowledge. These principles are reflected in internationally recognised guidance such as the Frascati Manual.
Importantly, R&D is not limited to traditional laboratory research. It may also arise in practical commercial environments where companies undertake systematic experimental or investigative work in areas such as engineering, software development, manufacturing, product development, process improvement, data science, artificial intelligence, materials development and other technology-driven fields.
In broad terms, the Department of Science, Technology and Innovation recognises qualifying R&D as encompassing activities aimed at discovering non-obvious scientific or technological knowledge, creating or developing inventions, functional designs, computer programs or related knowledge, making significant improvements to these outputs or conducting qualifying clinical trials.
The IT Buzz
The 2026 Budget makes it clear that Government is placing increasing emphasis on digital innovation as a national economic policy priority. As businesses accelerate their use of global reliance on data, automation and artificial intelligence, the infrastructure that supports these technologies is fast becoming central to competitiveness, productivity and future critical engines of economic growth. Recognising its importance, the South African government has elevated data infrastructure to the same priority level as traditional assets such as electricity, ports and transport networks.
The Minister’s outlined plans for new strategies in 2026 to boost investment in data centres and related infrastructure signals a clear intention. These efforts aim to position South Africa as a leading digital innovation hub on the African continent, supporting the expansion of advanced technologies and attracting international investment.
This creates an important opportunity for businesses investing in digital transformation R&D activity. Recent changes to the R&D Tax Incentive now allow claims for qualifying software developed for internal systems, including ERP platforms, fintech solutions, logistics engines and internal AI tools, which were previously excluded. For companies to build proprietary technology, this policy shift is expected to drive greater uptake of the commercial case for the incentive and encourage businesses to invest in innovation.
However, the opportunity should be approached with care. The incentive is generous, but it is also policy-driven and governed by specific legislative requirements. A project that is commercially important, costly or technically complex will not automatically qualify, unless it meets the statutory requirements for R&D.
This distinction is particularly relevant in the technology sector. Implementing an advanced platform, integrating existing tools or deploying AI into a business process may be complex and valuable, but complexity alone is not enough. The key question is whether the company is seeking to resolve a genuine scientific or technological uncertainty and whether the work has the potential to advance the underlying technology.
In computer science and information technology, there is often an assumption that if something is technically complex, it should automatically qualify. The short answer is no. One analogy I often use is that of a washing machine. Testing multiple detergents may involve experimentation to determine which one works best for colors, removes stains more effectively, or delivers better overall cleaning performances. However, simply comparing existing detergents is not necessarily advancing detergent technology.
AI can be viewed in much the same way. In our analogy, AI is the detergent. Unless your detergent is composed of nanocomponents that can cognitively tackle specific stains while being biodegradable, fix stitching and holes in clothing, actively restore fabric properties, adjust foaming based on water quality and machine capabilities, or dynamically releases cleaning agents in stages, there is no technological advancement (I exaggerate, but you get my point). Unless the work involves developing fundamentally new AI capabilities, rather than applying existing tools, there is likely no technological advancement. Likewise, there is no technological risk simply because a project may fail commercially or require significant investment; governments generally do not consider financial risk to be a technological uncertainty. As a result, the use of AI alone does not automatically make a project eligible for an R&D tax incentive.
In the AI context, simply using an existing AI tool is unlikely to be sufficient. By contrast, work aimed at developing new AI capabilities, overcoming technical limitations, improving model performance in a non-obvious way, or creating novel functionality may be more relevant. Financial risk, implementation risk or the risk that a product may not succeed commercially is not, on its own, the same as technological uncertainty.
A structured development process is therefore critical. Companies should be able to show a logical line of investigation, including the uncertainty being addressed, the alternatives considered, the testing performed and the results achieved. The emphasis should be on systematic testing rather than informal trial and error.
At the same time, businesses should not assume that only breakthrough inventions qualify. Many eligible projects involve incremental, but meaningful, technological advancement. The practical challenge is to identify the qualifying elements early and document the technical basis for the claim in a way that aligns with the legislative framework.
In a fast-moving digital environment, this assessment is becoming increasingly important. As more organisations invest in AI, automation, data platforms and internal software, the difference between applying existing technology and advancing technology will be central to determining whether the R&D Tax Incentive can be accessed.
Why businesses should prepare now
R&D incentive claims are strongest when qualifying activities are identified before or as they happen. Waiting until year-end or after project completion can make it difficult to reconstruct the technical uncertainty, experimental steps, people time, costs and outcomes needed to support the claim.
Companies should therefore review their innovation pipeline now, identify projects involving scientific or technological uncertainty, align technical and finance teams and put repeatable evidence-capture processes in place. This is especially important for sectors where R&D may not be labelled as “research”, such as software development, engineering, logistics, manufacturing, construction technology and applied AI.
How BDO can assist
BDO’s R&D Tax Incentives team assists businesses throughout the incentive lifecycle, from identifying potentially qualifying projects and assessing eligibility under Section 11D, to preparing technical documentation, quantifying qualifying expenditure, supporting applications to the Department of Science, Technology and Innovation and assisting with ongoing compliance and record-keeping.
We also help companies design practical processes to capture evidence of systematic investigation and maintain the financial records needed to support the claim. This allows technical and finance teams to approach the incentive proactively, rather than reactively.
If your business is investing in software, data, AI, engineering or other technology-led innovations, the R&D Tax Incentive may be worth considering. The BDO South Africa R&D Tax Incentives Team is available to help you assess the opportunity and prepare a robust, defensible claim.
South Africa’s Section 11D Research and Development (R&D) Tax Incentive offers businesses a valuable opportunity to reduce the after-tax cost of innovation. Where companies undertake qualifying scientific or technological development activities in South Africa, they may qualify for a 150% tax deduction on eligible R&D expenditure. This can include direct people costs, such as salaries and wages for employees involved in R&D, as well as qualifying materials, consumables, equipment usage and other inputs required to carry out the work.
To access the incentive, companies must obtain pre-approval from the Department of Science, Technology and Innovation and include the relevant approval with the annual SARS claim. With the programme now extended until 31 December 2033, businesses have a meaningful window to identify qualifying projects, strengthen their R&D documentation and embed robust processes before expenditure is claimed.
Definition of qualifying R&D
R&D tax incentives are used by many leading economies including Canada, the United States, the United Kingdom, Japan and Portugal. Although each jurisdiction applies its own legislative framework, the core policy objective is broadly consistent: to encourage businesses to invest in activities that advance scientific or technological knowledge. These principles are reflected in internationally recognised guidance such as the Frascati Manual.
Importantly, R&D is not limited to traditional laboratory research. It may also arise in practical commercial environments where companies undertake systematic experimental or investigative work in areas such as engineering, software development, manufacturing, product development, process improvement, data science, artificial intelligence, materials development and other technology-driven fields.
In broad terms, the Department of Science, Technology and Innovation recognises qualifying R&D as encompassing activities aimed at discovering non-obvious scientific or technological knowledge, creating or developing inventions, functional designs, computer programs or related knowledge, making significant improvements to these outputs or conducting qualifying clinical trials.
The IT Buzz
The 2026 Budget makes it clear that Government is placing increasing emphasis on digital innovation as a national economic policy priority. As businesses accelerate their use of global reliance on data, automation and artificial intelligence, the infrastructure that supports these technologies is fast becoming central to competitiveness, productivity and future critical engines of economic growth. Recognising its importance, the South African government has elevated data infrastructure to the same priority level as traditional assets such as electricity, ports and transport networks.
The Minister’s outlined plans for new strategies in 2026 to boost investment in data centres and related infrastructure signals a clear intention. These efforts aim to position South Africa as a leading digital innovation hub on the African continent, supporting the expansion of advanced technologies and attracting international investment.
This creates an important opportunity for businesses investing in digital transformation R&D activity. Recent changes to the R&D Tax Incentive now allow claims for qualifying software developed for internal systems, including ERP platforms, fintech solutions, logistics engines and internal AI tools, which were previously excluded. For companies to build proprietary technology, this policy shift is expected to drive greater uptake of the commercial case for the incentive and encourage businesses to invest in innovation.
However, the opportunity should be approached with care. The incentive is generous, but it is also policy-driven and governed by specific legislative requirements. A project that is commercially important, costly or technically complex will not automatically qualify, unless it meets the statutory requirements for R&D.
This distinction is particularly relevant in the technology sector. Implementing an advanced platform, integrating existing tools or deploying AI into a business process may be complex and valuable, but complexity alone is not enough. The key question is whether the company is seeking to resolve a genuine scientific or technological uncertainty and whether the work has the potential to advance the underlying technology.
In computer science and information technology, there is often an assumption that if something is technically complex, it should automatically qualify. The short answer is no. One analogy I often use is that of a washing machine. Testing multiple detergents may involve experimentation to determine which one works best for colors, removes stains more effectively, or delivers better overall cleaning performances. However, simply comparing existing detergents is not necessarily advancing detergent technology.
AI can be viewed in much the same way. In our analogy, AI is the detergent. Unless your detergent is composed of nanocomponents that can cognitively tackle specific stains while being biodegradable, fix stitching and holes in clothing, actively restore fabric properties, adjust foaming based on water quality and machine capabilities, or dynamically releases cleaning agents in stages, there is no technological advancement (I exaggerate, but you get my point). Unless the work involves developing fundamentally new AI capabilities, rather than applying existing tools, there is likely no technological advancement. Likewise, there is no technological risk simply because a project may fail commercially or require significant investment; governments generally do not consider financial risk to be a technological uncertainty. As a result, the use of AI alone does not automatically make a project eligible for an R&D tax incentive.
In the AI context, simply using an existing AI tool is unlikely to be sufficient. By contrast, work aimed at developing new AI capabilities, overcoming technical limitations, improving model performance in a non-obvious way, or creating novel functionality may be more relevant. Financial risk, implementation risk or the risk that a product may not succeed commercially is not, on its own, the same as technological uncertainty.
A structured development process is therefore critical. Companies should be able to show a logical line of investigation, including the uncertainty being addressed, the alternatives considered, the testing performed and the results achieved. The emphasis should be on systematic testing rather than informal trial and error.
At the same time, businesses should not assume that only breakthrough inventions qualify. Many eligible projects involve incremental, but meaningful, technological advancement. The practical challenge is to identify the qualifying elements early and document the technical basis for the claim in a way that aligns with the legislative framework.
In a fast-moving digital environment, this assessment is becoming increasingly important. As more organisations invest in AI, automation, data platforms and internal software, the difference between applying existing technology and advancing technology will be central to determining whether the R&D Tax Incentive can be accessed.
Why businesses should prepare now
R&D incentive claims are strongest when qualifying activities are identified before or as they happen. Waiting until year-end or after project completion can make it difficult to reconstruct the technical uncertainty, experimental steps, people time, costs and outcomes needed to support the claim.
Companies should therefore review their innovation pipeline now, identify projects involving scientific or technological uncertainty, align technical and finance teams and put repeatable evidence-capture processes in place. This is especially important for sectors where R&D may not be labelled as “research”, such as software development, engineering, logistics, manufacturing, construction technology and applied AI.
How BDO can assist
BDO’s R&D Tax Incentives team assists businesses throughout the incentive lifecycle, from identifying potentially qualifying projects and assessing eligibility under Section 11D, to preparing technical documentation, quantifying qualifying expenditure, supporting applications to the Department of Science, Technology and Innovation and assisting with ongoing compliance and record-keeping.
We also help companies design practical processes to capture evidence of systematic investigation and maintain the financial records needed to support the claim. This allows technical and finance teams to approach the incentive proactively, rather than reactively.
If your business is investing in software, data, AI, engineering or other technology-led innovations, the R&D Tax Incentive may be worth considering. The BDO South Africa R&D Tax Incentives Team is available to help you assess the opportunity and prepare a robust, defensible claim.