RDTI tobacco and gambling exclusions introduced into Parliament

Background  

The anticipated change to Australia’s approach to public investment in innovation has recently taken a step forward with the targeted tobacco and gambling exclusion from the R&D Tax Incentive regime being introduced into Parliament on 26 March 2025.  

Schedule 4 of the Treasury Laws Amendment (Delivering an Efficient and Trusted Tax System) Bill 2026 contains the proposed activities exclusion and reflects a shift away from industry‑neutral support for innovation to ensure that taxpayer innovation funding is aligned with broader public health and social policy objectives.

Following the announcement in the 2024–25 Mid‑Year Economic and Fiscal Outlook (MYEFO), Treasury released exposure draft legislation for consultation on 8 December 2025, closing on 30 January 2026.  

Although a commendable aim, RSM together with other stakeholders made detailed submissions critical of the potential breadth of the draft statutory definitions that will apply. Despite these submissions, the legislation introduced into Parliament is a replica of the exposure draft legislation, with nary a word changed.  

This article seeks to explore the nature of the exclusion and challenges that taxpayers will face on the retrospective enactment of these provisions from 1 July 2025.  

Overall observations  

R&D tax concessions have existed in Australian since 1985 to encourage innovation and technological advancement in all industries on an agnostic or sectorneutral basis, supporting any activity that meets the scientific and experimental statutory criteria. This approach matches most other advanced economies, with many adopting the principles laid out in the Frascati manual.  

To date, these sectorneutral innovation principles have globally avoided selective industry exclusion and the singling out of tobacco and gambling, both being regulated and lawful industries, does prima facie appear to be prejudicial given that other industries with significant publichealth impacts such as social media, alcohol, fossil fuels and ultraprocessed foods will all remain eligible for RDTI support.

Although the underlying concerns are valid, the question becomes whether regulations should be targeted to control or reduce the existing harmful behaviours, rather than seek to suppress underlying innovation given the risk of stifling broader innovation that could also have positive implications.  

Indeed, from an economic perspective, given the potentially extensive detrimental impacts on both software-based and advanced manufacturing activities, it is arguable that the exclusions will serve to weaken Australia’s competitiveness and reduce domestic R&D investment.  

With no other jurisdictions contemplating a similar approach, multinationals will seek to relocate valuable R&D activities to jurisdictions without punitive exclusions. By way of example, the 2018 Senate Inquiry into the R&D Tax Incentive heard that New Zealand was introducing a new RDTI program creating strong rivalry for investment in the ANZ region. Based on published data, in the five years to September 2024 total approved R&D expenditure in the New Zealand RDTI was in excess of NZD$7bn and rapidly growing. RSM Australia is witnessing Australian and multinational businesses locate R&D work to New Zealand. With the introduction of such industry specific and non-tax neutral measures, Australia runs the risk of whole industries to relocating their R&D teams and efforts across the Tasman.

In addition, pursuing these exclusions due to the public transparency information will set a precedential risk and open the door to future exclusions based on the prevailing political sentiment at that time rather than objective innovation criteria. These developments would significantly dilute the reputation of Australia as a place of objective innovation.  

Gambling-related RDTI exclusion  

The drafting of the proposed legislation in the Bill introduced into Parliament covers an exceptionally broad set of exclusions, with the statutory criteria likely to be insurmountable even for businesses operating adjacent to, but not wholly within the gambling industry.  

Since the exclusions are not appropriately qualified, the exclusion could be administered to apply in a manner not contemplated by the legislature. Recommendations to taper the broad exclusionary language to specify whether the linkage must be direct, primary, or incidental were not included in the final legislation.

For example, the current definition blanket-covers any type of activity carried out that “relates to” gambling and “gambling-like practices” (a term also used for the Digital Games Tax Offset (DGTO) legislation but not defined therein).

The wording of the proposed statute does also not make clear whether it is the purpose of the R&D activity, or the potential use of the output of the R&D results that is the key factor that will determine its eligibility. The breadth of the language used seems to imply that it would apply to both scenarios and that if a technology could be used in gambling, or is used by a gambling entity, it may fall under the exclusion even if the R&D itself has nothing or little to do with gambling. It is also unclear whether R&D on generic technologies that could potentially be later used by gambling operators is affected.

We believe that a more purpose-based or functional use test would have been more suitable where R&D activities should be excluded where the contemporaneous purpose of the activity (which must be defined under the RDTI regime) is specifically directed at gambling functionalities.  

The proposed huge breadth of terminology is already creating significant amounts of uncertainty for those businesses operating adjacent to, but not wholly within the gambling industry, as well as entities that develop mixed use technologies, where only part of the R&D activities may relate to elements of gambling-like functionality. For example, many technologies serve multiple industries – examples include:

• Random number generators used in gaming can also be used in cybersecurity.
• A behavioural analytics model can be used for both harm minimisation or for marketing.
• A payments engine could support gambling but also e commerce.

In addition, modern digital products often include mechanics that could be seen to resemble gambling, such as:

• Randomised rewards
• Loot boxes
• Probability based outcomes
• Gamified engagement loops

For example, the DGTO legislation excludes games with contain loot boxes, per the example in its Explanatory Memorandum. In practice, this has resulted in many fringe cases determining whether certain features in games meet the definition of a “loot box”. Such a broad exclusion has potentially excluded games otherwise eligible, and not strictly within the intent envisioned by the legislature. Given that the new RDTI legislation, inter alia, leverages the DGTO definition as one of three limbs for the exclusion, this will likely lead to a large portion of digital gaming / video games businesses being excluded from the RDTI. It is highly unlikely that this was an intended policy outcome, based on publicly released materials.

For other such affected adjacent entities, since the final legislation still adopts a blanket term of “gambling like practices”, it will remain ambiguous and unclear as to whether video game developers, digital engagement platforms, consumer apps and loyalty programs are unintentionally caught.

There is an important safeguard in the form of the carve out for those activities caught where they are ‘solely for harm minimisation”. However, the use of the explicit term “solely” means that in practice this carve out may only be construed and interpreted extremely narrowly.

Since “gambling related” remains broadly defined, it will also be unclear whether:

• a harm minimisation tool built within a gambling platform,
• a cross-industry behavioural analytics model, or
• a responsible gaming algorithm embedded in a wagering app

would represent activities that are “solely” for harm minimisation. It is also unclear whether harm minimisation R&D conducted by the gambling operators themselves is eligible, or only when conducted by third parties.

Regulatory guidance with useful examples would be welcomed, illustrating mixed-use and borderline exclusionary cases (e.g., payment systems, cybersecurity, fraud detection, generic platform infrastructure) as well as worked examples of eligible harm minimisation R&D (e.g., behavioural analytics, self-exclusion tools, real time risk detection).

Tobacco-related exclusion  

Many of the comments made above remain relevant to the proposed tobacco activities exclusion. The proposed legislative mechanism is also extremely broad with “Activities related to tobacco” being an undefined term and able to capture both upstream and downstream R&D that may be unrelated to the consumption of tobacco.

Such a broad exclusion is likely to have an impact on Australia’s advanced manufacturing capabilities with tobacco related R&D often involving capabilities that benefit the broader Australian and global economy including:

• Employment of materials science, chemists, engineers, data scientists, toxicologists.
• Usage of specialised research facilities such as labs and testing centres that also service other industries.
• Innovation surrounding supply chain capabilities including materials, packaging, biotech, and manufacturing.

As with the impact of the gambling exclusion on the software industry, the removal of RDTI support risks both job losses and capability erosion in the Australian advanced manufacturing industries.

Similarly, the ‘solely’ test for the tobacco related harm minimisation carve-out may also be insurmountable in practice since harm minimisation in this industry is understood to rarely be isolated. Most harm reduction research is integrated with broader product development, toxicology, materials science, and delivery mechanism research.

There could also be unintended outcomes such as reduced investment in safer alternatives with innovation in reduced risk products, such as heat not burn or noncombustible nicotine delivery systems, dependent on broader tobacco R&D programs that would be disincentivised and hence negate the government’s stated intention to support tobacco harm minimisation.

Conclusions

RSM consider that affected companies will face significant uncertainty, increased audit risks, and costly legal interpretation, even in the absence of evidence that these exclusions will serve to improve public health outcomes.

We also note that the draft legislation is proposed to be retroactive and will apply to income years starting on or after 1 July 2025 while many companies have multi-year R&D programs already underway with no grandfathering provision for projects that commenced before the announcement date.

Given the existing complexity of RDTI eligibility, it is to be hoped that there will be an early release of guidance and public rulings, updated RDTI guidance materials and an available channel of communication as to whether activities may fall within the exclusions. This would assist in reducing extended disputes and unnecessary compliance costs. 
 

FOR MORE INFORMATION

If you would like to learn more about the topics discussed in this article, please contact your local RSM office.