How CIOs Can Overcome the AI Trust Gap
CIO, Friday, October 30th, 2026
CIO's Mary Shacklett outlines five steps CIOs can take to build trust in AI as CEOs and boards grow wary of AI results.
In this CIO opinion piece, Mary Shacklett argues that trust in AI has become a concern for CEOs and boards, and that CIOs are expected to make AI pay off and are already blamed for missteps.
She recommends five steps: require quality data, models and business-linked metrics before launch; start with narrow, controllable projects; build a shutdown mechanism for every AI system and include it in disaster recovery plans; consider an outside audit for security and compliance; and vet AI vendors on quality, security and recovery.
She cites an MIT survey finding 95% of generative AI projects reported no return, and a Harris and Dataiku survey of 900 CEOs on job risk and explainability.