Submission 249
Employment Hero
Submitter group: Companies.
An AI model identified the submitter group and wrote the summary, key points and recommendation summaries. Counts do not measure importance or merit.
Summary
Read the original submission for the full text.
Employment Hero argues existing compliance burdens constrain SME AI adoption and new rules require gap analysis. It describes survey evidence favouring task redesign over job cuts among advanced adopters, while noting surveys are proxies and workers' discomfort matters. It proposes a sovereign frontier lab, superannuation-linked public benefits and creator compensation, and support for domestic domain-specific platforms including its own.
Key points
Quotes are the submitter’s own words. Text taken from documents, including scanned pages read by text-recognition software, can differ from the PDF. Check the original. Listed in the same order as on the submission’s page in the interactive guide.
It argues SME adoption is held back by the accumulated weight of existing compliance and regulatory cost rather than lack of appetite.
For the small and medium employers who make up the vast majority of the Australian economy, the barrier to deeper AI adoption is rarely appetite. It is the accumulated weight of existing compliance and regulatory cost that leaves owners with no time or capital left to invest in new technology.
It states that existing legislation is sufficient to manage AI-related risk and that further AI-specific compliance is not the answer to slow SME adoption.
Our position is that existing legislation is sufficient to manage AI-related risk and we do not believe the answer to slow SME adoption is a further layer of AI-specific compliance sitting alongside existing obligations.
Says its multinational survey found AI-core businesses reporting increased entry-level headcount more often than non-adopters (62% versus 30%).
Businesses with AI at the core of their operations report increasing entry-level headcount at more than double the rate of non-adopters, 62 per cent compared with 30 per cent, a finding from our AI Paradox at Work research covering more than 3,200 business leaders and 5,400 workers across Australia, New Zealand, the United Kingdom and Canada.
It says Australia has an opportunity to move from being a customer of overseas AI to being a builder and part owner of its own.
Australia has an opportunity to move from being a customer of overseas AI to being a builder and part owner of its own, while other countries are still working out their approach to frontier AI capability.
It argues businesses that hold the data, operate in a regulated environment and have customer trust are best placed to build valuable AI applications.
The businesses best placed to build valuable AI applications are the ones that already hold the data, already operate inside a regulated environment and already have the trust of the customers they would be serving.
It proposes that every Australian hold an equal stake in a sovereign AI entity or fund through superannuation-linked accounts.
Rather than treating sovereign AI as a conventional government asset, the proposal would give every Australian an equal, transparent stake in the resulting entity or fund, held through accounts integrated with the existing superannuation system, with returns flowing back as dividends, super contributions or tax credits
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Have the Office of AI catalyse a sovereign frontier lab through a global talent competition, faster researcher/engineer visas and seed compute/hosting funding, weighting business-building experience at least equally with academic/research distinction.
Under the proposal, the Office of AI would act as a catalyst, creating the conditions for the lab to succeed. This concept proposes running a global talent competition to attract researchers and engineers, supporting faster visa pathways for that talent and providing seed funding for the compute and hosting infrastructure a competitive lab needs to get off the ground. We would also encourage any such talent pathway to weigh people who have built and scaled technology businesses at least as heavily as academic or research distinction, since founders and operators are as important to a genuine national AI capability as researchers are.
Design any benefit-sharing from sovereign AI capability, including export revenue, so individual Australians hold a direct, transparent stake through the existing superannuation system rather than a new standalone entity or fund.
Design any benefit-sharing arrangement attached to sovereign AI capability, including revenue from AI Australia exports, so that individual Australians hold a direct, transparent stake through the existing superannuation system, rather than through a new standalone entity or fund structure.
Recognise the value of narrow, domain-specific AI models trained on Australian data in regulated industries and consider how to support Australian platforms building them, rather than treating sovereign AI solely as one national infrastructure project.
Recognise, alongside a frontier lab, the value of narrow, domain-specific AI models trained on Australian data inside regulated industries and consider how government can support Australian platforms already building this capability, rather than treating sovereign AI solely as a single national infrastructure project.
Before introducing AI-specific regulation, complete a gap analysis of existing legal frameworks (privacy, consumer, work health and safety, anti-discrimination) in consultation with industry to identify genuine gaps.
Before introducing AI-specific regulation, complete a gap analysis of existing legal frameworks, including privacy, consumer, work health and safety and anti-discrimination law, in consultation with industry, to identify where genuine gaps exist rather than assuming new AI-specific rules are the default response.
If new AI safeguards are found necessary, make them risk-based, outcomes-focused, technology-neutral and phased, with timeframes and compliance pathways reflecting small employers' resourcing constraints.
If new AI safeguards are found to be necessary, ensure they are risk-based, outcomes-focused, technology-neutral and phased, with implementation timeframes and compliance pathways that reflect the resourcing constraints of small employers rather than assuming enterprise-level compliance capability.
Treat the cumulative weight of existing regulatory and compliance obligations on SMEs, rather than the absence of AI-specific rules, as the primary barrier to adoption, and prioritise broad-based regulatory simplification alongside any AI-specific measures.
Treat the cumulative weight of existing regulatory and compliance obligations on small and medium businesses, rather than the absence of AI-specific rules, as the primary barrier to adoption identified in this inquiry, and prioritise broad-based regulatory simplification alongside any AI-specific measures.
A program corrected obvious text-recognition errors in this quote for display.
Extend existing SME training and upskilling incentives to cover practical, owner-focused AI literacy, including tool selection and safe implementation, delivered where possible through digital employment platforms SMEs already use.
Extend existing SME training and upskilling incentives to cover practical, owner-focused AI literacy, including how to select tools appropriate to a small business and how to implement them safely, delivered where possible through the digital employment platforms SMEs already use.
A program corrected obvious text-recognition errors in this quote for display.
Consider whether government has a role in pointing small and medium employers toward existing, tested AI adoption frameworks, as a low-cost way of narrowing the adoption gap.
We would encourage the Committee to consider whether government has a role in pointing small and medium employers toward existing, tested adoption frameworks like this, as a low-cost way of narrowing the adoption gap described in section 4.
Draw on real-time, aggregated payroll and employment data held by registered digital service providers, plus periodic surveys, to track how AI adoption changes hiring, wages and task composition, and revisit settings as evidence develops.
Draw on real-time, aggregated payroll and employment data already held by registered digital service providers, alongside periodic survey instruments, to track how AI adoption is actually changing hiring, wages and task composition as this inquiry's recommendations are implemented and revisit settings as that evidence develops.
A program corrected obvious text-recognition errors in this quote for display.
Test directly, rather than assume, whether deeper AI adoption means fewer jobs.
Much of the public conversation about AI and work proceeds from an assumption that deeper adoption means fewer jobs. Our data does not support that assumption and we think the Committee should test it directly rather than accept it by default.
Take Australian workers' discomfort about AI seriously as a feature of workplace adoption.
Our AI Paradox research found that 45 per cent of Australian workers are concerned that AI is making them more replaceable, and 42 per cent say using AI to do parts of their job feels like cheating, even when it is measurably improving their output. Australian workers appear to feel this more acutely than counterparts in the other markets we surveyed. We urge the Committee to take this discomfort seriously as a real feature of how AI adoption is landing inside workplaces.
Judge Australia's sovereign AI strategy by whether it maximises human employment and opportunity, not simply economic output.
We think Australia's sovereign AI strategy should ultimately be judged by the same measure we hold ourselves to: whether it maximises human employment and human opportunity, not simply economic output.
If government establishes an initiative offering SMEs real-world AI adoption examples, include Employment Hero's experience, including its recruitment outcomes.
If government or industry stands up any initiative to give small and medium businesses credible, real-world examples of AI adoption to work from, we would welcome Employment Hero's experience, including the recruitment outcomes described above, being one of the examples such an initiative points to.
Give Australian authors, artists, news organisations and other data owners shares or royalties when they choose to contribute work to training sovereign models.
Australian authors, artists, news organisations and other data owners who choose to contribute their work to training sovereign models would receive shares or royalties in return, turning the copyright and training data debate into a shared incentive rather than a standing conflict.
Read the original
In some quotes a program corrected obvious text-recognition errors for display.