CCI staff share recent surveys, reports and analysis on risk, compliance, governance, infosec and leadership issues. Share details of your survey with us: editor@corporatecomplianceinsights.com.
C-suite more confident than practitioners about AI’s accuracy
More than a quarter of business executives (26%) say audits have detected AI-generated mistakes in materials that have made their way to the board or external audiences, according to a survey from corporate reporting platform Workiva.
The survey of more than 2,200 global finance, risk and sustainability professionals also highlighted a modest split between the C-suite and practitioners when it comes to AI’s trustworthiness. About 4 in 5 executives (84%) said they’d be at least somewhat confident in AI output appearing in an annual report without human review, while only 76% of practitioners said the same.
Investors are tracking both AI’s accuracy and its profitability. Most institutional investors in the survey (89%) said they were concerned about AI accuracy in corporate disclosures, while 51% said they are tracking revenue growth to gauge AI’s return on investment.
Other findings include:
- 47% said their organizations formally model the climate-related impacts of AI investments and include them in the business case before approval.
- Among executives at public companies or those on the path to IPOs, more than 90% were at least somewhat likely to say they would continue to publish quarterly earnings figures even if the SEC adopts a rule to permit semiannual reporting.
- 88% of executives said training their workforce on AI is a top priority in the second half of the year.
Less than 30% of organizations have full AI governance
Organizations are putting money into AI governance, but few have fully operationalized controls, according to a survey by Schellman, an IT compliance and cybersecurity firm.
Surveying 525 US-based AI governance professionals, Schellman found that almost all (90%) said their organizations have allocated funding for AI governance, but only 27% say their AI governance is mature, operational and continuously monitored.
Despite funding AI governance, only two-thirds (64%) of organizations have a baseline formal, documented AI acceptable use policy that’s communicated to employees, respondents reported. Less than half (44%) have documented AI-specific incident response procedures, while 57% maintain a formal AI governance policy.
Organizations’ nascent AI governance is being outpaced by agentic AI adoption, according to the survey. A vast majority (86%) have put agents into testing and 46% have agents in production. Organizations don’t always require human oversight of AI agents with 38% reporting they require human review only for high-risk or high-impact AI agent decisions, 32% requiring human review for all agent actions and 22% having defined thresholds that would trigger a human review.
Data center boom means risk is booming, too
As data center project investment is expected to balloon to more than a trillion dollars by 2027, risks for these projects are also skyrocketing with climate and labor being some of the prominent liabilities, according to a report by business insurance provider Allianz Commercial.
Nearly 80% of global data center capacity is located in areas exposed to heightened natural catastrophe risk, the report said. Flood, wildfire and wind exposure is highest in the Americas, affecting 86% of capacity, while chronic heat and drought stress is greatest in Asia Pacific, where 89% of capacity is exposed. Natural catastrophes rank second behind fire as the cause of the highest financial loss for data centers, according to the report. Willful acts, including physical and cyber crime, are the third-highest cause.
The data center boom is being driven by the need for computing power for AI, which is being adopted rapidly by enterprises, quickly outpacing their governance of the technology, according to surveys.
The post 26% of Execs Say Audit Has Caught Public-Facing AI Mistake appeared first on Corporate Compliance Insights.
