Strong foundations and human adoption will decide who gets real value from AI, says Commonwealth Superannuation CTIO

AI conversations are getting noisier, but the real divide is still familiar.

Some organisations are strengthening core systems, simplifying complexity, and building the discipline to use new tools well.

Others are layering AI over weak architecture, fragmented data, and unclear ownership, then wondering why the results feel shallow.

In this ADAPT Insider conversation, Andrew Matuszczak, Chief Information and Transformation Officer at Commonwealth Superannuation Corporation, argues that AI will not rescue poor foundations or weak execution.

For him, the more useful question is whether the business has done enough core transformation work to make AI worth adding in the first place.

Listen to the full episode on Apple Podcasts, Spotify, and YouTube.

Key takeaways:

• Fix the core systems first, because AI will scale weakness just as quickly as it scales value.
• Treat AI adoption as a people challenge as much as a technology challenge, with practical training and use cases that make the value real.
• Use first principles, simple governance, and disciplined investment choices to avoid hype driven decisions.

AI only becomes valuable when the foundations can carry it

Andrew is blunt on this point.

If the core systems are still monolithic, overly complex, and poorly integrated, AI will only amplify those weaknesses.

That is why CSC has spent years doing the less glamorous work first, upgrading core platforms, improving its data stack, and simplifying how systems connect before trying to push AI deeper into customer and employee workflows.

That approach matters because AI is being asked to sit on top of operational reality, not a demo environment.

If the foundations are weak, the result is cosmetic change rather than meaningful improvement.

If the foundations are strong, AI can start improving areas such as call centre interactions, knowledge access, and workflow support in ways that actually hold up.

The harder problem is getting people to use it well

Andrew also argues that many organisations are still treating AI as a technology rollout when the bigger challenge is human adoption.

Fear, uncertainty, and doubt still shape how people respond, from boards asking if they are being left behind, to employees wondering what the tools mean for their role.

That is why CSC has focused heavily on practical, role based uplift rather than abstract AI messaging.

Short training, relatable use cases, and internal champions are all part of building confidence.

The aim is to help people understand how these tools improve the way they work, not just what the tools are called.

In Andrew’s view, adoption will move faster when people can connect AI to everyday value instead of seeing it as another layer of technical change.

Simplicity and discipline matter more than speed

Andrew is also sceptical of the generic way AI is discussed.

In his view, calling everything AI hides the real question, which is what problem the business is actually trying to solve.

Different tools suit different needs, and that makes first principles thinking more useful than broad enthusiasm.

That same discipline applies to governance and investment.

Boards are pushing management teams to move faster, while also asking harder questions about risk, control, and exposure.

The organisations that handle that tension best will be the ones that keep governance practical, make clear choices about where value sits, and avoid turning AI into another expensive layer of technical debt.

Speed alone will not decide the winners. Clarity, simplicity, and execution will.