In his seminal treatise on generative AI, Bob Dylan warned us that “the times, they are a’changing”.
The reality of this prescient warning is, based on our client base, that licensees, advisers, and compliance teams are gaining momentum as they start to explore the practical value of generative AI for their businesses.
I’m not suggesting for a moment that this interest is limited to the leading advice businesses that comprise our client base. The rise of Generative AI (GenAI) represents a paradigm shift for financial services, promising new efficiencies and challenges.
This projection underscores the transformative economic potential of AI technologies and the importance of addressing governance gaps to harness their full benefits. While many advisers, compliancersTM, and licensees are cautious about adopting new technologies, others see these tools as a competitive advantage. What is undisputable from our perspective is that as the debate around GenAI continues, it’s crucial to explore its potential applications within compliance and advice while addressing the inherent risks.
Below are six general principles we’ve developed to guide financial advisers, licensees, and compliance teams in integrating GenAI effectively and responsibly.
1. Treat GenAI as a Competent but Inexperienced Assistant
Before your demoralised staff jam their sabots into your mainframe to protest their inevitable replacement, take a moment to appreciate that GenAI should complement—not replace— people, compliance frameworks and human judgment. For instance, even where we use integrated GenAI to pre-screen submissions for compliance with precedents, the law or ASIC guidelines, we consider that step as an input, not the final output. Where the AI flags potential issues, real advice compliance experts provide the final assessment, ensuring that nuanced client needs and ethical considerations are addressed. We know how important this is; too many clients relied on inaccurate AI SoA Reviews because of the failure to consider context, perspective and nuance.
A 2019 article about AI adoption from Fast Company emphasised this crucial truth: AI is not coming to replace human expertise entirely but to change the nature of work.
Tools like ChatGPT or enterprise-grade AI platforms can streamline processes, but their real value lies in enhancing, not supplanting, professional expertise. For example, GenAI can help summarise compliance reports, draft client communications, or identify regulatory updates. Human review is essential to ensure these outputs are accurate, appropriate, and fit for purpose.
In practice, treat GenAI as a well-intentioned but inexperienced team member. Invest time, manage and mentor the model and train it in your business and philosophy. You wouldn’t allow the new graduate in your office to work autonomously and entirely unsupervised, so take the same approach with your new GenAI tools.
AI allows professionals to focus on the more nuanced, creative, and strategic elements of their roles — tasks that machines cannot yet replicate.
The Fast Company article highlights the irreplaceable value of human expertise and judgment in financial services, even as GenAI transforms operational efficiencies. Similarly, GenAI can serve as a valuable assistant in financial services, but professionals must always maintain accountability for its outputs.
Like a new graduate going on vibes, GenAI’s eagerness to help means that many models are prone to “hallucinating”; offering plausible but false responses or manufacturing evidence. In financial services, this could lead to non-compliant or misleading content. Compliance teams must implement robust review processes and ensure outputs are consistent with regulatory requirements like those set out by ASIC in Regulatory Guide 256 (Client Review and Remediation Conduct) and Regulatory Guide 271 (Internal Dispute Resolution).
ASIC’s 2024 report warns that emerging governance gaps in AI adoption could exacerbate risks, with 61% of licensees planning to increase AI use in the next 12 months.
In a recent article in the National Law Review, Daniel Knight from K&L Gates highlighted that Australian financial services licensees must ensure AI aligns with governance and ethical frameworks. This vital point underscores the critical need for licensees to integrate AI responsibly while mitigating potential risks.
ASIC’s 2024 speech reminds businesses that new technologies do not absolve them of their responsibilities toward good governance and consumer protection. Oversight is critical regardless of technological maturity.
Remember, while the tool and medium might be new, your legal obligations haven’t changed.
ASIC has repeatedly reiterated that existing regulations already apply to AI use. Its recent enforcement actions, such as the case involving insurance pricing models, demonstrate the importance of complying with current laws.
This is a crucial point echoed by the Australian Institute of Company Directors (AICD), emphasising the need to integrate AI-specific governance frameworks alongside existing regulations to address evolving regulatory expectations effectively.
Under these circumstances, it is essential to implement and maintain proactive governance to effectively manage the rapid adoption of AI technologies and mitigate associated risks.
2. Mentor and Direct GenAI
Success with GenAI doesn’t hinge on quirky prompts but on precise, task-specific inputs.
Some elaborate prompt engineering is available on TikTok and other places, but in my experience, clarity and specificity will get you most of the way. Precision and clarity are essential for leveraging AI for compliance and financial services roles regardless of your chosen model. Unless Licensees and Compliance staff encode industry best practices into their prompts, their outputs will not meet professional and regulatory standards. For example, when asking GenAI to generate a strategy review, the input should include clear parameters such as client goals, risk tolerance, and relevant regulatory considerations.
We’d all probably agree that providing GenAI with relevant, concise context leads to better results. You should also appreciate that overloading systems with unnecessary data—whether in drafting advice or generating reports—risks inaccuracies and inefficiencies.
Clarity and focus are as important in AI inputs as in client communication. One practical approach is to train staff on prompt engineering techniques to maximise the value of GenAI tools.
3. Balance Enthusiasm and Resistance
Not all organisations or clients are equally ready for AI adoption.
ASIC’s 2024 report highlights the emergence of governance gaps in licensees’ AI adoption, underscoring the importance of ensuring robust oversight and clear accountability structures during implementation.
This is why the Australian Institute of Company Directors (AICD) recommends that boards include AI-related expertise to strengthen accountability and ensure effective governance of AI initiatives.
While some licensees may eagerly embrace these tools, others may proceed cautiously. Tailoring AI integrations to align with the technological maturity of your team and client base is essential. For example, a larger institution may have the resources to experiment with cutting-edge AI solutions, whereas smaller practices might prefer proven, cost-effective tools.
GenAI isn’t ready to produce compliant Statements of Advice, but advisers using GenAI for any other purpose must also be transparent about the origin of their tools and recommendations. If GenAI contributes to the advice you produce, this should be disclosed, ensuring clients and compliance teams can distinguish between human expertise and machine assistance. Transparency builds trust and reinforces accountability.
4. Protect Data and Privacy Rights
Australian financial services operate under stringent data protection laws.
ASIC’s Report 798 highlighted that only 12 licensees had policies addressing fairness or inclusivity, underscoring the need for robust data governance frameworks to ensure equitable and compliant AI adoption.
This striking gap demonstrates the pressing need for financial services providers to address ethical considerations as they expand their use of AI technologies. Remember, ASIC’s Report on AI adoption identified a range of data risks, such as AI tools improperly handling sensitive information.
This aligns with stakeholders’ emphasis on transparency and accountability, recognising that the lack of proper governance can lead to significant risks in AI adoption.
AI governance needs rigorous oversight, and one should not ignore the broader challenges in data governance, particularly in ensuring compliance and maintaining public trust in AI systems. Two points to consider carefully are: First, GenAI must not infringe on a party’s copyright or moral rights. Second, any use of GenAI must align with the Privacy Act 1988 (Cth) and other relevant legislation, ensuring client information is secure and data sovereignty is preserved.
This means ensuring that your AI providers comply with Australian data storage requirements and do not retain sensitive client data for training their models. The challenge of this requirement is one of the reasons we recommend that users do not submit personal, commercial, or proprietary information to a GenAI model.
When integrating GenAI, ensure compliance with intellectual property laws and licensing agreements to avoid unauthorised use of proprietary content. Without permission, uploading third-party financial models to GenAI systems could breach licensing terms, exposing your practice to legal risks.
5. Leverage GenAI for Efficiency, Not Shortcuts
GenAI’s most significant benefit is its ability to automate routine tasks, freeing up advisers’ time to focus on higher-value activities.
For example, ASIC’s Report 798 highlighted the transformative applications of AI in fraud detection and call analysis. These real-world cases demonstrate how automation improves operational efficiency and enhances the quality of oversight in financial services.
This core idea underscores how AI can transform the financial services sector by improving operational efficiency while allowing professionals to focus on strategic tasks. For example, AI can draft initial compliance reports, monitor regulatory updates, or generate tailored client communications. However, shortcuts must never come at the expense of human oversight. The AI-generated visuals for LinkedIn posts should demonstrate why every GenAI output must be verified for accuracy and relevance before implementation.
Routine and repetition are critical because GenAI tools are most effective as part of a continuous improvement process. Regular feedback loops can help refine AI-generated outputs over time. For instance, if a GenAI model consistently produces suboptimal compliance summaries, reviewing and adjusting the prompts or underlying data can improve results. This iterative approach ensures the technology evolves alongside your professional and business development.
6. Build a Compliance Framework for AI Use
Before implementing GenAI, licensees should develop a comprehensive compliance framework tailored to AI applications.
ASIC recommends that licensees (depending on their nature, scale and complexity) establish an executive-level AI governance committee with defined responsibility and authority over AI-related risks and oversight.
This committee should provide regular reports to the board or a designated committee to ensure accountability and alignment with organisational objectives. This framework should include guidelines on acceptable use, data management protocols, and responsibilities for reviewing AI-generated outputs.
The Australian Institute of Company Directors (AICD) also recommends embedding AI risk management into broader enterprise governance frameworks to ensure comprehensive oversight.
But it doesn’t end there. To optimise AI deployment, ASIC advises businesses to align AI governance strategies with their organisational objectives and incorporate the eight Australian AI Ethics Principles into their strategy. For example, a clear policy could outline when GenAI may be used for client-facing communications versus internal processes.
Summary of the Australian AI Ethics Principles
The Australian AI Ethics Principles provide a framework for the responsible development and use of AI technologies:
- Human, Social, and Environmental Wellbeing: Ensure AI benefits society, individuals, and the environment.
- Human-Centred Values: Respect human rights and diversity while supporting individual autonomy.
- Fairness: Promote inclusivity and prevent discrimination.
- Privacy and Security: Safeguard data and respect privacy.
- Reliability and Safety: Ensure AI systems operate safely and reliably.
- Transparency and Explainability: Make AI decisions understandable and transparent.
- Accountability: Establish clear mechanisms for accountability.
- Contestability: Enable individuals to challenge and resolve AI-driven decisions.
Integrating AI is neither a silver bullet nor a set-and-forget strategy. GenAI models and agents must be continually monitored to ensure they deliver consistent and accurate results.
ASIC highlights the necessity of regularly reviewing and updating governance arrangements to keep pace with evolving AI risks and adoption levels. Without proactive adjustments, licensees risk creating governance gaps that may harm consumers.
Successful GenAI integration also requires a workforce that understands its capabilities and limitations. Investing in AI literacy training can empower advisers and compliance teams to use these tools effectively. Training sessions could cover prompt engineering, data privacy considerations, and the ethical implications of using AI in financial advice.
Regular evaluations can identify areas for improvement and ensure the technology remains aligned with organisational goals.
The Australian Institute of Company Directors (AICD) emphasises the importance of dynamic governance practices that adapt to the rapidly changing AI landscape, ensuring long-term alignment with strategic objectives.
GenAI introduces ethical questions related to transparency, accountability, and fairness. ASIC highlights the need for sufficient human and technological resources to support ethical AI deployment and ensure organisations have the skills and infrastructure required to address these challenges effectively.
The Australian Institute of Company Directors (AICD) echoes this point. It emphasises the importance of ongoing board-level education and engagement with AI to maintain informed oversight and align technological initiatives with strategic goals.
ASIC’s Report 798 observed that few licensees tested for algorithmic bias, highlighting a significant gap in ensuring AI-generated outputs align with the principles of fairness and equity. The solution, robust testing for algorithmic bias to prevent discrimination or unfair outcomes, seems reasonable but requires auditing to ensure that AI models and agents don’t inadvertently reinforce biases or exclude specific demographics from receiving fair treatment.
Treat the GenAI model or agent you use as an Outsourced Service Provider and manage it in a manner consistent with your other service providers. Consider establishing key performance indicators (KPIs) for AI applications so you can track their effectiveness over time. For instance, you might measure the accuracy of AI-generated compliance reports or the time saved on routine tasks to assess the appropriateness of your approach.
Managing Output Risk
GenAI offers transformative potential, but it’s not a panacea.
The Compliance Catechism you’ve most consistently heard emphasises balancing innovation and accountability in financial services. Only by grounding GenAI use in established compliance principles can financial advisers and licensees unlock efficiencies while meeting their professional and regulatory obligations.
For instance, consider using GenAI to enhance client engagement by generating tailored market updates or automating routine compliance tasks to reduce administrative burdens. These applications improve efficiency and allow advisers to dedicate more time to building client relationships and providing strategic insights.
As the financial services industry navigates this new technological frontier, the same proactive approach you’ve taken to compliance needs to be applied to GenAI adoption. It’s not just a matter of regulatory compliance because without adhering to these principles, advisers and licensees can’t embrace innovation without sacrificing the trust and integrity that underpin their success.
If you’re considering how AI might enhance your compliance or advice processes, Assured Support offers tailored, actionable solutions to navigate this technological shift confidently.
If you liked this, we recommend:
High Stakes, Clear Rules: ASIC’s 2025 Agenda to Safeguard Consumers and Markets
Governance Essentials for AFS Licensees: A Practical Guide
Avoid the AI Trap: How to Stay Compliant and Ahead of the Game
Guide for Advisers and Licensees: AI and Financial Services
FAQ Section
1. What is GenAI, and how can it benefit financial services?
Generative AI (GenAI) refers to advanced AI systems capable of creating content, analysing data, and automating tasks. In financial services, it can streamline compliance processes, draft client communications, detect fraud, and more. By automating routine tasks, GenAI allows professionals to focus on strategic and creative aspects of their roles, enhancing efficiency and service quality.
2. What are the risks of using GenAI in financial services?
While GenAI offers significant efficiencies, it also carries risks such as data breaches, algorithmic bias, and the creation of inaccurate outputs (“hallucinations”). Financial service providers must implement strong governance frameworks to mitigate these risks, comply with regulations like ASIC’s guidelines, and ensure human oversight of AI-generated content.
3. How should financial advisers integrate GenAI into their workflows?
Financial advisers should use GenAI as an assistant rather than a replacement for human expertise. Start by clearly defining the AI’s tasks, such as pre-screening investment strategies or summarising compliance reports. Always review AI-generated outputs for accuracy, align them with regulatory requirements, and disclose the AI’s role to clients to maintain trust and transparency.
4. What are the ethical considerations when using GenAI in financial services?
Ethical use of GenAI involves ensuring fairness, avoiding discrimination, safeguarding data privacy, and maintaining accountability. GenAI also has significant environmental impacts that need to be considered. Adopting AI governance frameworks that align with principles like the Australian AI Ethics Principles can help address these challenges while fostering trust in AI-driven processes.
5. How can financial services ensure compliance with GenAI use?
Compliance requires a comprehensive AI governance framework that includes regular oversight, transparency in AI operations, and robust data protection measures. ASIC recommends creating an executive-level AI governance committee to oversee risks and align AI use with regulatory and ethical standards. Regular reviews and updates ensure compliance as technology evolves.
