As the uptake of generative artificial intelligence (GenAI) technologies accelerates across Australian businesses, practical and legal questions arise around the management of intellectual property (IP) rights. The purpose of this article is to highlight the emerging issues, explain how current Australian law applies, and outline what businesses should be doing now to protect their IP assets.
Why is this topic important now?
The Australian IP regime remains technology-neutral, meaning the traditional IP frameworks still apply even though the underlying technologies (such as GenAI) are evolving rapidly.
Recent commentary from IP Australia and policy-makers emphasises that GenAI poses novel issues for inventorship, authorship, and data use.
Organisations using or deploying GenAI tools — whether to develop new products, automate creative content, or augment design and software workflows — face IP risks if those issues are not addressed.
Failure to proactively manage IP risks in the GenAI context may lead to uncertainty, unintended loss of rights, exposure to infringement claims, or difficulties in commercialising AI-derived outputs.
Key IP issues to consider
1. Ownership, inventorship and authorship of AI-generated output
Under Australian patent law, an “inventor” must be a natural person. This restricts the ability to list an AI system itself as inventor, which may in turn affect the registrability of inventions like patents, trade marks, plant breeder rights, circuit layout rights or designs generated with GenAI. It is also likely to affect the ability to enforce copyright or common law trade mark rights against infringement. For copyright protection, protection is typically afforded to works that exhibit sufficient human authorship (skill, labour, judgment, intellectual effort). Pure machine-generated output may fall outside that protection.
Because of this, businesses using GenAI should ensure robust documentation of the human contribution to inventions or works (e.g., who drafted or selected inputs, what parameters were designed, who supervised the output, who made revisions, etc).
Contractually, the rights in GenAI-generated output should be clearly assigned or licensed between parties (developers, users, adopters, vendors).
2. Use of third-party AI tools and licensing of output
Many GenAI tools rely on datasets, models or components developed by third parties — it’s critical to review licences of those tools to understand who owns the output, what rights are granted, and who bears the risk of underlying third-party claims.
If your business is licensing GenAI services (for example, content creation, design generation, code synthesis), you should assess whether the service provider has cleared rights to all incorporated works/data and whether any restrictions apply to commercial use.
Consider whether your AI system itself might be protectable (as software, design, trade mark, trade secret) and whether vendor contracts facilitate your retention of rights or control of resultant innovations.
3. Trade marks, designs and branding in a GenAI world
When GenAI is used to generate logos, product designs, brand assets, or user-facing content, questions arise about originality, distinctiveness and ownership.
Before proceeding to register a trade mark or design based on GenAI output, review whether the output is unique, whether third-party data has been used, and whether the human contribution is sufficient to warrant registration.
From a brand-risk perspective, promoting GenAI-generated content without addressing IP clearances may expose the business to reputation or infringement risks.
4. Training data, copyright and data-use issues
Many GenAI models are trained on large datasets that may include copyrighted text, images or other works. While no special “AI copyright exception” exists yet in Australia (it does in other jurisdictions), the question of whether TDM (text-and-data-mining) or other uses are covered under current copyright exceptions is under active debate.
Businesses that use GenAI should assess whether their training data and use cases are licensed or permissible, and whether the output is derived from or substantially replicates any protected work. Our prediction is that GenAI software owners will continue to strike commercial agreements with authors of creative works to be granted a licence to use their creative work as training data. But we can’t rule out litigation against businesses that use the outputs generated by GenAI for commercial advantage.
Evidence and documentation of data provenance, processing, and transformation are becoming commercially and legally important.
5. Strategic preparation for legal/regulatory change
Australia’s regulatory responses to AI (including from the Australian Competition & Consumer Commission and other bodies) are evolving, and future developments may impact IP, liability and governance obligations.
Although IP law has not changed specifically for GenAI, businesses that plan now — auditing their AI usage, clarifying IP ownership, and updating contracts — will be better positioned to respond quickly to change.
From a strategic viewpoint, identifying IP assets, managing risk and aligning AI deployment with commercial objectives is a competitive advantage.
Practical steps for your business
Conduct an IP audit:
Map your use of GenAI technologies, identify the IP you are creating (software, models, output, data), and clarify who owns or licenses each relevant component. Note that many existing software products are wrapping GenAI suite products into their platforms, sometimes in ways that are not obvious to the software user.
Our IP +T team provides a service to help businesses identify their critical, and valuable, IP and understand the true ownership of ideas when created using Generative AI or other software in the creation process.
Review and update contracts:
Ensure your agreements with employees, contractors, vendors, and service providers allocate IP rights in GenAI developments and output, require appropriate warranties and clearances, and address third-party data use.
Implement governance and documentation:
Establish policies that track data sources, model training, human involvement (inputs, supervision, editing), and output usage. Maintain records to support authorship/inventorship decisions.
Evaluate IP protection pathways:
Determine whether your GenAI-derived innovations are protectable by patents, designs, trade marks or whether other protections (trade secret, contract) are more appropriate given the human-contribution and ownership issues.
Stay informed and agile:
Monitor developments in AI regulation, IP case law (particularly around AI-generated works), and technological change. Adopt flexible strategies to respond to shifts in the legal landscape.
Train your staff:
Invest in educating your staff about GenAI and these issues. The more they know, and the more awareness they have around these complexities, the more likely they will do the right thing to protect your business.
Why choose us?
The Intellectual Property + Technology team at Clifford Gouldson Lawyers combines deep legal expertise with practical business acumen. We assist clients to protect and manage IP assets, draft and negotiate technology and IP contracts, and advise on the evolving interplay between innovation, AI and IP rights. Whether your business is developing its own GenAI systems, incorporating third-party tools, or commercialising AI-derived output, we can help you navigate the issues and develop the right strategy.
If you’d like to discuss how your business can manage IP risks and opportunities in the GenAI era, please contact us.
For further information contact Ben Gouldson.
Dismissed by ChatGPT: The decent thing to do?