Fear of mistakes slows down AI adoption

Despite the growing attention for artificial intelligence (AI), there remains significant hesitation among finance professionals. Over half (53%) say that the fear of making mistakes outweighs the fear of missing out when it comes to AI. A quarter (24%) admit they only adopt AI applications once the majority has done so. These findings come from research conducted by surveying 359 employees working in the financial sector. 

A clear majority (62%) are not yet ready to hand over full control to an AI-driven model. Expectations for technology are high: 46% believe that while human colleagues are allowed to make mistakes, AI is not. Furthermore, 38% place more trust in the output of their colleagues than in that of AI systems. 

The primary concern is privacy and data security, cited by 59% of respondents as the greatest risk. Other major concerns include security threats (47%) and the fear of losing human oversight (38%). These concerns contribute to the fact that 21% believe the risks of AI outweigh its potential benefits. 

While there is clear hesitation, there is also optimism. 51% of respondents view AI as an opportunity rather than a threat. Still, there is a sense that AI has yet to live up to expectations. For instance, 18% believe AI is overhyped and that many applications are not feasible in practice. Additionally, 16% feel that the environmental costs of AI outweigh its benefits. 

Adil Bohoudi, Managing Director of Conclusion AI 360, comments: “Organizations are balancing between enthusiasm and caution when it comes to AI. Trust in technology is steadily increasing, but the threshold remains high when risks or errors are involved. Companies, especially in the financial sector, prefer to wait until the technology can be implemented flawlessly rather than lead the way in this transition. However, this fear is unnecessary—as long as businesses integrate AI step by step and in a responsible manner. Establish clear frameworks, ensure transparency, and maintain control over data and processes. That way, you lay a solid foundation for safe and successful AI adoption.”