Oceania lags global AI trust benchmarks, SAS finds

SAS has published research showing Oceania remains behind global benchmarks for trustworthy AI, even as adoption rises across the region.

Research conducted by IDC for SAS found the share of organisations in Oceania running integrated or transformative AI programmes increased to 48.7% in 2026 from 38% a year earlier. Trust in AI, however, has not advanced at the same pace, limiting returns on those investments.

Organisations with the strongest governance, data quality and auditability practices reported at least double the return on investment from AI deployments compared with peers. By contrast, fewer than one in 20 organisations classed as trustworthy AI laggards reported the same level of return.

Global data pointed to a much wider divide. Organisations investing in trustworthy AI measures were 15 times more likely to report strong or high returns from AI projects, with 62% doing so compared with 4% of those that did not prioritise those practices.

Oceania's trustworthiness index rose 3.3 points to 59.5, but the region still trailed global benchmarks. The study also recorded gains across all five dimensions used to assess trustworthiness.

Those dimensions were data quality and governance; model governance and oversight; explainability and fairness; responsible AI policy; and audit and accountability. In the study, a trustworthy AI leader was defined as having an average score of 80 or higher.

Trust gap

One of the clearest constraints on value in Oceania is a gap between rising AI use and weaker organisational readiness in oversight, infrastructure and workforce skills. The report argued that better governance alone would not be enough to reverse a 13-point decline in the region's Impact Index.

The researchers also identified an infrastructure gap: the absence of locally hosted advanced large language models used by organisations in Oceania. As a result, businesses routinely send sensitive data overseas for processing, creating what the report described as a sovereignty gap and a structural limit on growth.

The study suggested local verification and grounding systems are the most practical response for now, while broader infrastructure investment will be needed to close the gap. No single organisation, it added, can solve that issue alone.

IDC and SAS also examined why employees override AI recommendations, treating it as a sign of weak trust in automated decisions. The data showed 29.9% cited insufficient explanation, while 22.7% pointed to factual errors and the same share cited bias or fairness concerns.

Those manual interventions can erode productivity and profitability. Stronger data foundations, better explainability and tighter bias detection would help reduce override rates and improve the value organisations derive from AI-based decision-making, the report found.

Skills decline

While AI use is spreading, some measures tied to organisational preparedness moved in the opposite direction. The share of organisations hiring or training staff for AI ethics, compliance and risk management fell 30.8 percentage points to 26.9%.

Regular audits and impact assessments also declined, dropping 9.1 percentage points to 27.5%. Those falls came even as the region's overall trustworthiness score improved and the gap with global peers narrowed.

The findings suggest workforce readiness may become a more pressing issue as organisations try to move from experimentation to broader deployment. Companies that invest in training alongside technology are seeing stronger returns because they understand when, how and why to apply AI, according to the report.

Industry comparisons in the wider global study showed sectors are moving at different speeds. In banking, 85% of AI leader banks had established governance frameworks, compared with 29% of laggards.

Among public sector organisations, 41% of leaders were increasing investment in trustworthy AI by more than 20%. In life sciences, 23% of organisations had scaled AI across the company, the highest share among the industries covered.

The survey drew on responses from 2,699 decision-makers involved in data and AI initiatives across 28 countries. It covered banking, insurance, life sciences and the public sector.

"When AI works, it's incredibly impactful," said Bryan Harris, Chief Technology Officer, SAS. "However, it is well documented that state-of-the-art agents can have error rates that exceed 25% on complex tasks[1] - which is unacceptable in high-stakes decision-making. In order to achieve accuracy and repeatability, organisations must embed domain expertise into agentic workflows, while keeping people at the centre of governance and oversight. Organisations that do this successfully will close the trust gap and gain a competitive advantage in the market with AI."

"As AI becomes more autonomous, organisations face a new challenge: maintaining confidence in systems people don't fully understand," said Chris Marshall, Vice President, IDC. "Our findings show that stronger oversight, explainability, accountability and data foundations are becoming prerequisites for scaling AI successfully."

For Oceania, the region now faces a choice between treating current results as a base for further work or pausing after a more cautious year, SAS said. "While more and more organisations in the Oceanic region are adopting 'Integrated' or 'Transformative' AI programs, rising from 38% to 48.7%, we have observed that the sector's underlying trust in AI hasn't kept pace with the market's ambitions", said Craig Jennings, Regional Vice President & Managing Director, Asia Pacific, SAS. "As a result, organisations haven't been able to sink their teeth into AI, creating a tight, credible trust gap between Oceania. While markets like Australia have faced regulatory considerations and others continue to navigate workforce skills or decisioning challenges, there is a strong opportunity to build AI value across the region. However, its near-term trajectory will depend on whether the market treats 2026 as a floor to build from or a correction to wait out".