# User trust and accuracy challenges in AI

> Live situation record from CLSTR: https://clstr.news/situations/user-trust-and-accuracy-challenges-in-ai
> Updated: 2026-09-24T02:13:40.000Z. Sources: 4. Developments: 2.

Reports indicate significant challenges regarding user trust and accuracy in artificial intelligence applications.

A study by SAS and IDC found that 97.2% of users override or correct AI recommendations, a behavior attributed to a lack of explainability. The report noted that organizations prioritizing trustworthy AI practices, such as data governance and auditability, are 15 times more likely to report a high return on investment than those that do not.

Complementing these findings, a survey by Knowledge Holdings Co., Ltd. reported that approximately 80% of users have encountered inaccurate responses, including hallucinations and incorrect product specifications. Despite these errors, user retention remains high, with 85% of respondents intending to continue using AI services, though some plan to proceed with increased caution.

## Timeline

### 2026-09-24: Knowledge Holdings survey finds 80% of AI users encounter errors

A survey of 306 AI users shows that 80% have encountered incorrect information, yet 85% intend to continue using AI tools despite errors like hallucinations and incorrect product data.

2 sources. https://clstr.news/cluster/knowledge-holdings-survey-finds-80-of-ai-users-encounter-errors

### 2026-09-15: SAS and IDC report: 97% of users override AI recommendations

A SAS and IDC study finds that 97% of users correct AI recommendations due to poor explainability, while trustworthy AI practices increase the likelihood of high ROI by 15 times.

2 sources. https://clstr.news/cluster/sas-and-idc-report-97-of-users-override-ai-recommendations

---
Cite as: User trust and accuracy challenges in AI. CLSTR, https://clstr.news/situations/user-trust-and-accuracy-challenges-in-ai