Trustworthy AI Delivers 15x Stronger ROI, SAS-IDC Study Finds

سبتمبر 15, 2026 ENGLISH اضف تعليق

Organizations that want to turn artificial intelligence investments into measurable business returns may need to focus on more than deployment speed and technological capabilities. A new report from SAS, featuring research insights from IDC, finds that trustworthy AI practices are emerging as a decisive factor separating organizations that generate strong returns from those struggling to realize value.

Based on a global survey of 2,699 business and technology decision-makers across 28 countries, the second annual Data and AI Impact Report: The New Economics of Trust found that organizations applying trustworthy AI practices were 15 times more likely to report strong ROI from their AI projects.

The findings show that organizations with stronger governance, higher data quality and greater auditability consistently outperform their peers. Although these organizations represent a relatively small segment of the market, they report at least twice the ROI from AI deployments, while fewer than one in 20 trustworthy AI laggards report comparable results.

Egypt’s AI Ambitions Put Trust in Focus

The findings are particularly relevant to Egypt, which ranked first in Africa and 51st globally among 195 countries in Oxford Insights’ 2025 Government AI Readiness Index, climbing 14 places in a single year.

The ranking reflects the momentum behind Egypt’s National AI Strategy 2025–2030, which aims to increase AI’s contribution to GDP to 7.7% and train 30,000 AI specialists by 2030.

As Egypt expands its AI ambitions across government and industry, the SAS report highlights a critical question: whether this growth can translate into measurable returns. According to the findings, success will depend not only on deploying AI faster, but also on implementing systems that employees and regulators can trust, understand and hold accountable.

Employees Are Overriding AI They Cannot Understand

One of the report’s central findings is that employees are increasingly reluctant to rely on AI systems when they cannot determine whether an output is correct or understand how the system reached its final decision.

As AI becomes more autonomous, the inability to explain decisions becomes an increasingly significant business liability. The report therefore identifies explainability as a critical component of successful AI adoption.

A lack of confidence in AI decisions can also lead employees to override recommendations and make manual corrections. While such interventions may provide an immediate safeguard, they can widen the AI trust gap and create additional costs in terms of employee time, productivity and profitability.

Because AI decisions are ultimately dependent on the quality of the underlying data, organizations seeking to reduce AI override rates must also strengthen their data foundations.

97.2% of Users Override AI Recommendations

The research reveals that 97.2% of users override AI-generated recommendations in at least some situations.

The leading reason employees choose to override AI is not necessarily that the recommendation is incorrect. Instead, the primary trigger is the system’s inability to explain the reasoning behind its decision.

The study also found that trust declines as AI systems become more autonomous. Confidence falls from 76% for generative AI to 66% for agentic AI, highlighting growing concerns about systems that increasingly operate with less direct human intervention.

Trustworthy AI Creates a Major ROI Gap

The report identifies a widening divide between organizations that prioritize trustworthy AI and those that do not.

Importantly, the organizations gaining the greatest value from AI are not necessarily using fundamentally different technologies. Their advantage comes from how they govern, manage and oversee those technologies.

Organizations investing in trustworthy AI measures are 15 times more likely to report strong or high ROI, with 62% reporting such results compared with only 4% among organizations that do not apply comparable practices.

Organizations with the strongest trustworthy AI practices also achieve 1.85 times greater gains across 13 different business outcomes, including revenue growth, cost savings and improved customer experience.

The performance gap could continue to widen, as 85% of AI leaders with trustworthy practices plan to increase their investment in this area by more than 10% during the current year.

Weak Data Foundations Are Holding AI Back

Despite growing AI adoption, many organizations continue to deploy AI on severely underdeveloped or outdated data infrastructure.

Without a strong data foundation capable of supporting transparency and explainability, organizations face greater difficulty governing AI effectively and extracting the expected business value from their investments.

The study found that only 17.5% of enterprises have a fully optimized data infrastructure mature enough to meet the demands of agentic AI, with weaknesses in data infrastructure negatively affecting performance.

Organizations with an optimized data foundation are four times more likely to expect strong ROI from AI projects. They are also six times more likely to mandate the data-quality and explainability controls required to establish trust.

Banking Leaders Turn AI Governance Into a Competitive Advantage

The research covers four key industries: banking, insurance, life sciences and the public sector, based on responses from decision-makers with knowledge of or influence over their organizations’ data and AI initiatives.

Within banking, AI leaders are moving beyond regulatory compliance and treating robust AI governance as both a competitive advantage and an operational necessity.

The findings show that 85% of AI-leading banks have established governance frameworks, compared with just 29% of lagging organizations.

Public Sector Investment in Trustworthy AI Accelerates

The public sector is also preparing to increase its focus on trustworthy AI.

According to the report, 41% of public sector leaders plan to increase their investment in trustworthy AI by more than 20% over the year ahead.

This rate of planned investment growth is comparable to that of the most ambitious organizations across all industries surveyed.

Life Sciences Leads Enterprise-Wide AI Expansion

Life sciences organizations are showing strong progress in scaling AI across their businesses.

The report found that 23% of life sciences organizations have scaled AI company-wide, representing the highest proportion among the industries examined in the study.

The finding highlights the growing importance of moving beyond isolated AI projects toward broader enterprise adoption supported by appropriate governance, data infrastructure and accountability.

Human Oversight Remains Essential

“When AI works, it’s incredibly impactful,” said Bryan Harris, CTO at SAS.

Harris noted that state-of-the-art agents can have error rates exceeding 25% on complex tasks, a level he described as unacceptable for high-stakes decision-making.

To achieve greater accuracy and repeatability, organizations need to embed domain expertise into agentic workflows while keeping people at the center of governance and oversight.

Organizations that successfully strike this balance can close the trust gap and establish a competitive advantage as AI adoption accelerates.

Explainability and Accountability Are Becoming Prerequisites

“As AI becomes more autonomous, organizations face a new challenge: maintaining confidence in systems people don’t fully understand,” said Chris Marshall, Vice President at IDC.

Marshall said the findings demonstrate that stronger oversight, explainability, accountability and robust data foundations are becoming prerequisites for organizations seeking to scale AI successfully.

The shift toward increasingly autonomous AI therefore places greater emphasis on the ability of organizations to understand, monitor and govern the decisions produced by these systems.

What Makes AI Trustworthy?

The report defines trustworthy AI as artificial intelligence designed to be reliable, fair, secure, compliant with regulatory standards and capable of clearly demonstrating how it reached a decision.

Users and decision-makers at every level of an organization must be able to hold an AI system against a predetermined chain of accountability when its output is incorrect or incomplete.

Trustworthy AI also requires effective governance and evidence that systems comply with clearly defined rules and standards.

Five Criteria Define Trustworthy AI Leaders

As part of the study, organizations were scored on a 100-point scale across five dimensions of trustworthy AI.

Organizations achieving an average total score of 80 points or higher were classified as trustworthy AI leaders.

The five evaluation criteria are:

Data quality and governance

Model governance and oversight

Explainability and fairness

Responsible AI policy

Audit and accountability

The Business Case for Trustworthy AI

The findings from SAS and IDC point to a broader shift in the economics of AI. Deploying advanced systems is only one part of the equation; organizations also need the governance, data quality, explainability and human oversight required to make those systems dependable.

For businesses and governments expanding their AI strategies, building trust is increasingly becoming a practical business requirement rather than simply an ethical or regulatory consideration.

The full study and report are available through SAS at sas.com/ai-impact.