A new SAS report finds that 41.1% of UAE respondents cite lack of explanation as the leading reason for overriding AI recommendations.
As organizations adopt increasingly autonomous AI systems, employees are still stepping in to override decisions they do not understand, according to a new report from SAS, with research insights from IDC.
The second annual Data and AI Impact Report: The New Economics of Trust found that 97.2% of users override AI-generated recommendations in at least some cases. The most common reason, regardless of whether the AI output was considered correct, was that the system could not explain how it reached its decision.
The finding is particularly relevant in the UAE, where 41.1% of respondents identified insufficient explanation as the leading reason for overriding AI recommendations.
The report also found that trust declines as AI systems become more autonomous. Trust stood at 76% for generative AI, compared with 66% for agentic AI.
UAE sees sharp rise in AI trustworthiness
The UAE recorded one of the fastest gains in the study, with its Trustworthiness Index rising 30.2 points to 65.7 in 2026, putting it above the global benchmark.
The index measures organizations across five areas: data quality and governance, model governance and oversight, explainability and fairness, responsible AI policy, and audit and accountability.
Michel Ghorayeb, Managing Director of SAS UAE, said the increase reflects the rapid move from AI ambition to adoption in the country, pointing to the UAE AI Strategy and Dubai AI Roadmap as drivers of governance deployment across government and the private sector.
However, the report also identifies a gap between priorities and implementation. Data quality and governance has emerged as the leading reliability priority among UAE organizations, at 62.2% in 2026. Yet only 13.5% mandate data-quality processes for every AI project.
Trust and returns are linked
Globally, the study found a significant difference in reported returns between organizations with strong trustworthy AI practices and those without them.
Organizations investing in trustworthy AI measures were 15 times more likely to report strong or high ROI from their AI projects, with 62% reporting strong or high ROI compared with 4% among other organizations.
Organizations with the strongest trustworthy AI practices also reported 1.85 times greater gains across 13 business outcomes, including revenue growth, cost savings and customer experience.
The report found that 85% of organizations with these practices are increasing their investment in the area by more than 10% this year.
Data infrastructure remains another challenge. Only 17.5% of enterprises have a fully optimized data infrastructure considered mature enough for the demands of agentic AI.
Organizations with an optimized data foundation were four times more likely to expect strong ROI from AI projects and six times more likely to mandate data-quality and explainability controls.
The findings are based on a survey of 2,699 decision-makers across 28 countries, covering banking, insurance, life sciences and the public sector.
