From AI FOMO to Outcomes: Reflections from the Digital.NSW Showcase

[ikrieger]

Earlier this week I hosted a roundtable at the Digital.NSW Showcase at Randwick Racecourse in Sydney. It brought together senior executives from across NSW Government – people who are right in the thick of designing and delivering digital services for the state.

In my role as Innovation Officer at ServiceNow, I wasn’t there to talk about platforms or products. The goal was simpler, and harder: get past the AI buzzwords and have an honest conversation about what’s actually working, what isn’t, and what needs to change if we want AI to deliver real outcomes for citizens.

---

AI beyond FOMO

I opened with a simple question:

Given the way Gen Z and Gen Alpha already use AI in their daily lives, how is government thinking about its own adoption and deployment?

The responses were candid.

A lot of early AI work in government has been driven by a fear of missing out. Agencies picked up multiple pilots and proofs-of-concept just to “see what AI could do” – without always being clear about the problem they were solving or the outcome they wanted to achieve.

That experimentation phase wasn’t useless, but its limits are now obvious. It’s very easy to end up with a shelf full of pilots and few improvement in the experience of a citizen trying to get a service.

The tone in the room has shifted. Leaders are now more willing to say:

• “We’re not going to turn on AI just for the sake of it.”
• “If we can’t link this to a departmental outcome or a citizen outcome, we shouldn’t be doing it.”

In other words, AI is maturing from shiny object to delivery tool. The hard work now is having the uncomfortable conversations about saying no to “AI for AI’s sake” and yes to a smaller number of well-governed, outcome-focused initiatives. The proof of this was in the announcements and showcases at the Digital.NSW event.

---

Workforce and skilling: meeting talent where it is

We then turned to the workforce.

The talent market the public sector is competing in has changed. It’s not just about attracting “digital natives” straight out of university. It’s also about:

• People looking for flexible, portfolio-style careers, splitting their week between public service and a side gig.
• People returning to the workforce after a break (for example, parents coming back from parental leave).
• Potentially even late-career professionals who still want to contribute, but not in a traditional full-time role.

The conversation quickly moved from “How do we hire more people?” to “How do we redesign roles, pathways, and learning so people can actually say yes to us?”

One of the more encouraging signals was a growing openness to micro-certifications and on-the-job learning. A senior exec talked about shifting away from treating a traditional degree as the only entry ticket. Instead, they’re starting to think in terms of:

• Smaller credentials that can be earned over time.
• Practical experience gained while working.
• Skills that can be built and updated as technology changes, rather than locked in at graduation.

I shared my own career story, moving from my chosen profession into technology in the ’90s without a formal tech degree and it landed. That kind of non-linear path is becoming normal again, and government can either design for it or miss out on a huge pool of capable people.

We also touched on some very unglamorous, but critical, capability gaps. One example raised was radio frequency management – a highly specialised area where a lot of expertise is ageing out of the system. Some agencies are turning up to university open days, in their own time, showing students that these “boring but essential” jobs are worth considering.

AI might help with parts of these roles over time, but it can’t replace deep domain understanding overnight. If anything, it raises the bar on how intentional we need to be about skilling and reskilling.

---

Governance: from frameworks to real guardrails

Inevitably, the discussion came back to governance.

Government already has a lot of the machinery in place – risk councils, investment bodies, assurance processes. No one is starting from scratch. But there’s still a gap between having governance on paper and having it function as real guardrails for AI projects.

One useful insight from a previous roundtable with government was that not everyone is fully aware these frameworks exist, let alone feels confident navigating them. From the centre, it’s easy to assume “we’ve communicated this”; on the edge, it often looks like a maze.

At this roundtable, the mood was more self-aware. Leaders recognised that:

• Government does not need to be a first-mover on AI. In fact, there’s often more value in being a thoughtful second-mover, learning from others’ mistakes and focusing on safe, stable implementation.
• The challenge is less about inventing new AI governance frameworks, and more about plugging AI into the existing ones in a way that is clear and usable.

We’ve seen this movie before with cybersecurity. At first, it was treated as a technical bolt-on. Over time, it became embedded in risk, procurement, design, and delivery. AI now needs the same treatment: something that is baked into decision-making, not bolted onto it.

---

Where this leaves us

Walking away from the lunch, a few things stood out for me:

• The FOMO phase is ending. There’s a clear desire to move from “try everything” to “prioritise what actually shifts citizen outcomes.”
• Workforce conversations are finally catching up with reality. Micro-credentials, flexible careers, and non-linear pathways aren’t edge cases; they’re the new normal.
• Governance isn’t the blocker – awareness is. The structures are mostly there. The work now is making them visible, accessible, and confidently used across all levels of government.

For all the complexity, this is a good place to be. It’s a sign that AI in government is starting to grow up.

The next step is to stay disciplined: keep asking “What outcome are we trying to shift?”, invest in people as seriously as we invest in platforms, and treat governance as an enabler rather than an afterthought.

That’s where the real value – and the real trust – will come from.