#USA #Artificial Intelligence #CARY #Trustworthy AI #SAS Research

Organizations Winning with Trustworthy AI

According to the latest findings from SAS, organizations that adhere to trustworthy AI practices have a significantly enhanced chance of seeing strong returns on their AI investments. The research conducted by IDC for SAS highlights that these organizations are a staggering 15 times more likely to report high ROI from their AI initiatives compared to their counterparts who do not prioritize trustworthiness in AI.

The Importance of Trustworthy AI

Trustworthy AI refers to systems designed to be fair, reliable, and secure—the kind that can also clearly articulate the reasoning behind their decisions. A recent study finds that 97.2% of users have overridden AI-generated recommendations at least once. The primary reason for these overrides? A lack of explanation from AI models regarding their outputs. This phenomenon indicates a significant gap in trust that organizations must bridge to ensure effective AI deployment. Therefore, as AI technologies increase in autonomy, the demand for explainability and accountability becomes even more pressing for organizations.

Bryan Harris, the Chief Technology Officer at SAS, emphasizes the notion that when AI performs optimally, its impact can be groundbreaking. However, he warns about the risks of high error rates in complex tasks, which can reach up to 25%. Maintaining human oversight and embedding domain expertise into workflows is essential to mitigate these risks and foster trust in AI processes.

Risks of Ignoring AI Trustworthiness

The challenges do not end at user trust; they extend to organizational productivity and profitability as well. Many employees, feeling uneasy about AI decisions, revert to manual overrides, perpetuating inefficiencies. This behavior leads not only to lost productivity but also inhibits the organization's potential for AI-driven insights. The fact that trust levels drop from 76% for generative AI to 66% for agentic AI underlines the growing concern regarding AI's capability as it becomes increasingly autonomous.

ROI Disparity and Investment in Trustworthy AI

SAS's research reveals a significant ROI disparity between organizations that are committed to using trustworthy AI practices and those that are not. Major findings from the report include:

• - Organizations prioritizing trustworthy AI are 15 times more likely to report strong ROI from AI projects, with a 62% success rate compared to only 4% for lagging organizations.
• - Companies exhibiting the best practices in trustworthy AI enjoy up to 1.85 times greater gains across various business metrics, such as growth in revenue, reduced costs, and improved customer experience.
• - A notable 85% of AI leaders committed to trustworthy practices plan to increase their investments by over 10% this year.

These insights indicate that businesses optimizing their AI strategies for trustworthiness will not only outpace their competition but will also achieve greater operational efficiency.

The Data Foundation Challenge

A significant issue facing many companies today is the underdevelopment of their data infrastructure. The SAS report indicates that only 17.5% of enterprises possess a fully optimized data setup that meets the demands of agentic AI. This lack of foundation adversely affects performance and undermines efforts towards transparency and explainability in AI governance. Organizations equipped with a robust data foundation are found to be four times more likely to foresee strong ROI from AI projects. Furthermore, they are six times more inclined to prioritize data quality necessary for building trust.

Key Findings and Industry Trends

The report, which surveyed 2,699 decision-makers across 28 countries in four key sectors (banking, insurance, life sciences, and public services), emphasizes several industry-specific dynamics:

• - In banking, proactive governance regarding AI is becoming a competitive necessity, with 85% of leading banking firms establishing robust governance frameworks, compared to a mere 29% among laggards.
• - Public sector entities are on track to scale trustworthy AI investments by 20% or more in the upcoming year, reflecting dedication to advancing responsible AI practices.
• - In the life sciences sector, 23% of organizations have effectively implemented AI at a company-wide scale—the highest adoption rate seen across all studied industries.

Conclusion

To summarize, the SAS report elucidates the importance of trustworthy AI in realizing substantial ROI from AI investments. Organizations must adopt strong governance, enhance data quality, and ensure explainability in AI systems to successfully navigate the evolving landscape of artificial intelligence. The benefits of these measures not only improve trust but also solidify a competitive advantage in the marketplace. For more detailed insights, consider exploring the full report at sas.com/ai-impact.