It’s important to separate the hype from the reality.

• Automation refers to the use of technology to execute predefined, repetitive tasks—typically using workflow tools, scripts, or robotic process automation (RPA). It is deterministic, rules-based, and highly applicable to the operational activities common in superannuation and investment management.
• Artificial Intelligence, by contrast, involves systems that can learn, interpret patterns, and adapt over time. AI may be valuable in predicting member behaviour, interpreting unstructured data, or identifying investment anomalies—but it requires data maturity and robust governance foundations.

Too often, firms look to AI for transformation without first addressing the inefficient, manual processes that hold back their core operations.

Most investment firms and superannuation funds in Australia are still heavily reliant on spreadsheets, siloed platforms, and manual controls. Key business functions such as trade reconciliations, member registry updates, unit pricing, and corporate actions are often:

• Labour-intensive
• Inconsistent across platforms
• Dependent on key personnel
• Prone to error and delay

Rather than jumping ahead to AI, these organisations would gain significantly more by automating these known, repetitive processes—reducing cost, improving accuracy, and laying the groundwork for future innovation.

A. Lower Cost, Faster ROI

Automation initiatives—particularly those that target repeatable back- and middle-office tasks—can be implemented quickly, often without requiring system overhauls or specialist skills. Compared to AI, they carry fewer risks and deliver value faster.

B. Supports Regulatory Compliance

Regulatory obligations under APRA's CPS 230 (Operational Resilience) and SPS 530 (Valuation Governance) require enhanced control environments. Automation directly supports compliance by reducing human error and improving auditability.

C. Builds Data Readiness for AI

One of the biggest barriers to AI adoption is poor data quality. Automating data capture and validation processes improves data integrity—making AI possible later, when the organisation is ready.

Simply put: if your business processes are still manual, introducing AI will only amplify the chaos.

There are no doubt exciting use cases for AI in this sector - predictive modelling for member outcomes, intelligent fraud detection, natural language processing for fund documents, and more.

However, these use cases require clean, governed, and accessible data, as well as internal capability to monitor, validate, and explain machine learning outputs. Without those building blocks- most of which depend on automated workflows - AI initiatives are likely to fail or stall.

To ensure meaningful and sustainable digital transformation, firms should adopt a phased approach:

• Step 1 – Automate: Target high-volume, rules-based processes in operations, compliance, and reporting. Start with quick wins.
• Step 2 – Standardise Data: Centralise data sources, apply validation rules, and establish data ownership and governance.
• Step 3 – Introduce AI Selectively: Only after the above are in place should firms explore AI-driven use cases, ideally in sandbox environments with measurable business impact.