Inaction may be the biggest AI risk

Thursday, 23 July 2026

As Australian businesses race to turn artificial intelligence (AI) into a source of reinvention and growth, inaction is emerging as a significant handbrake that prevents organisations from moving beyond experimentation to execution.

Despite the rapid pace of AI advancement, many firms remain stuck in the test and trial phase, limiting their ability to turn pilots into profits. 

But according to PwC Australia Assurance partner, and the firm’s 2025 AI Innovator of the Year, David Ma, the biggest bottleneck to AI-led transformation is not moving at all.

Speaking on PwC Australia's podcast series on responsible AI, Ma said the first obstacle some companies encounter is hesitation, with others treating early setbacks as evidence that AI does not work.

"It's a journey, it's new and sometimes, that can be scary. But the change is happening at a rapid pace, so the sooner you start, the better,” Ma said.

The urgency around AI deployment is a growing reality for Australian businesses. PwC's 29th Global CEO Survey found the number one concern for local executives is keeping up with the pace of technological change. Leaders who lag could miss outsized returns compared to more decisive and ‘AI fit’ competitors, who generate revenue and efficiency gains that are up to 7.2 times higher.

Strong AI starts with robust data foundations

For Ma, AI-powered innovation begins by building trusted and fit-for-purpose foundations, with high-quality data and governance frameworks as the key starting points. 

Proactively embedding trust-by-design from the outset will allow businesses to fully capture the 'AI trust dividend' in the long-term, with risk management moving from a perceived compliance constraint into an ROI enabler.

While the infrastructure can require significant investment, Ma argues this approach enables more advanced applications over time and makes AI sustainable, scalable, and secure.

"Time and time again, data comes up as the biggest challenge individuals and organisations have," he said.

Based on PwC’s 2026 AI performance study, Australia is leading in this space, with 73% of local companies saying the apply robust, up-to-date protections for data, models and underlying architecture. The top performers are 1.7 times more likely to use a documented responsible AI framework.

AI is a momentum game

Many businesses may delay implementation while searching for the perfect use case, but Ma believes that approach often creates unnecessary indecision. Instead, he’s encouraged organisations to identify practical opportunities where AI can deliver tangible value quickly.

"Go for the low-hanging fruit first. Start off with the easy wins and get some runs on the board," Ma said.

Like most transformative technologies, successful execution is rarely linear. 

83% of those surveyed in PwC’s 2026 AI performance study said they’ve participated in AI pilots within their organisation, but it took an average of 6.8 months after the pilot launch for it to begin providing value. While the impacts may not be immediately felt, those that persist are better positioned to ultimately embed AI at scale.

Responsible AI needs people at the centre

While Ma is optimistic about what AI can deliver for businesses and individuals, he also insists that innovation cannot come at the expense of continued human oversight across the AI ecosystem.

PwC’s 2026 AI Jobs Barometer points to growing demand for people who can apply human judgement, manage risk, and take accountability as organisations operationalise AI. New tasks in the most AI-exposed occupations are 2.5 times more likely to require trusted human skills such as critical thought, creativity, and empathy. 

When asked what human involvement should look like, Ma compared responsible AI use to other autonomous technologies which ultimately require people making decisions and taking control.

"Traditionally, we drive our cars, and that’s human in the loop. We have our hands on the steering wheel, put our foot on the pedals, and we steer where it goes. On the other end, it's humans out of the loop, self-driving, fully autonomous cars.” 

“What responsible AI looks like is striking the right balance. At PwC we call this human at the helm."

Innovation requires a mindset shift

The pace of change is forcing workers and companies to continuously adapt. Employees in the most AI-exposed occupations have developed an average of 187 new skills since 2019 according to PwC’s 2026 AI Jobs Barometer, as roles evolve and new ways of working emerge.

With technology moving quickly, Ma argued it can be difficult for business leaders and individuals to rethink established behaviours and assumptions, which often blocks AI-fuelled growth.

Intention also defines innovation, with leading organisations strategically using AI to reinvent work, overhaul operations, and enter new growth markets, rather than find efficiencies and reduce costs.

You can watch the full podcast episode here.