AI investment is rising, but productivity isn’t.

Matt Boone, Senior Research Director at ADAPT, reframes the problem facing government leaders: the gap isn’t about technology adoption, but about the ability to convert investment into measurable, organisation-wide outcomes.

The productivity gap is an execution failure, not an innovation problem

Australia doesn’t lack innovation; it lacks the ability to scale and operationalise it into measurable productivity gains.

Continued investment without clear returns erodes executive trust, especially when CFOs already believe a significant portion of technology spend is wasted. Without translating AI into tangible outcomes, future funding and momentum are at risk.

Matt highlights that only 6% of agencies report clear, embedded productivity gains, with most seeing value only in isolated pockets. At the same time, CFOs estimate ~40% of tech spend is wasted, creating a credibility gap between investment and impact.

Complexity is increasing faster than capability

In trying to empower workers with more tools, organisations are unintentionally increasing friction and reducing productivity.

Tool sprawl and legacy environments compound inefficiency, making it harder for employees to realise the benefits of AI. This creates a paradox where investment in productivity tools undermines productivity itself.

Knowledge workers now juggle ~10 enterprise applications on average (up from six), while 40% of agencies are layering new technology onto old processes, effectively automating inefficiency rather than removing it.

Scaling AI is a readiness problem, not a technology problem

The biggest blockers to AI impact sit between “buying” and “doing”, in workforce, governance, and operating models.

Without readiness, organisations accelerate risk without gaining value. Investments in advanced capabilities like agentic AI outpace the ability to manage, govern, and integrate them safely, increasing the likelihood of failure at scale.

According to ADAPT data, only ~9% of agencies are scaling agentic AI, yet 73% plan to invest in it. Meanwhile, no leaders report high confidence in AI governance, and 75% of data and AI leaders say they are not prepared to manage an agentic workforce.

AI won’t close the productivity gap on its own; only disciplined execution, readiness, and shared accountability will.

Key takeaways:

• The productivity gap exists because organisations are investing faster than they can operationalise value.
• Workforce readiness and governance, not technology, are the real constraints to scaling AI. Treat workforce readiness and governance as core delivery work, not supporting activities.
• Demonstrable, early wins are essential to building trust and unlocking further investment. Reduce complexity by redesigning processes, not layering new tools on old systems.