Investors want to see measurable results from artificial intelligence (AI) investments. At the same time, evidence of AI-generated errors and poor data quality has created a high level of distrust in corporate disclosures.
Most investors (89%) are concerned about the accuracy of AI-generated information in the disclosures — with almost half (47%) saying they actively look for such errors, according to a global survey from Workiva, The Verification Gap: AI Enthusiasm Runs Into Reality.
Investor concerns are shared, to a lesser extent, by the companies. While 84% of executives and 76% of practitioners are at least somewhat confident in the accuracy of AI outputs appearing in annual reports without human review, 26% of executives said that internal audits of AI use detected errors that had reached investors, external audiences, or company boards.
The survey gathered responses in May from more than 2,000 finance, legal, risk, and sustainability professionals, including more than 800 C-suite executives, from organisations in North America, Latin America, Europe, and Asia-Pacific. The survey also polled more than 300 institutional investors in the UK, US, and Canada.
Despite accuracy concerns, the pressure is rising for organisations to produce tangible results from AI. Investors in the survey reported tracking several metrics to gauge AI’s ROI: revenue growth rates (51%), sales conversion rates (42%), internal rate of return on AI projects (40%), time savings reported by management (40%), and percentage of tasks done without human intervention (40%).
Nearly half of executives (48%) are optimistic that their companies will see measurable returns on AI investments within the next 12 months, primarily from revenue growth (58%) and time savings (47%).
The data suggests some executives struggle to separate enthusiasm from reality, with 71% reporting that “poor data quality has at least moderately impacted the use of AI in financial and sustainability reporting”, the survey said. Twenty-seven per cent said poor data has significantly blocked deployment in key workflows, and just 11% said their organisation’s data quality is sufficient for AI use.
“The good news for organisations who have been investing in data governance all along is that they may have a head start in unleashing AI on their data — and uncovering faster paths to growth,” the survey said. “For everyone else, the data gap has consequences that have investors on alert.”
— To comment on this article or to suggest an idea for another article, contact Steph Brown at Stephanie.Brown@aicpa-cima.com.