Experts highlight that despite the rapid adoption of AI agents by businesses, significant doubts about their accuracy could lead to future risks. Autonomy in decision-making is crucial for agents to deliver value, yet experts remain wary of these ‘black boxes’ that obscure the reasons behind specific actions. The transparency issue poses challenges for scaling agents as autonomous tools. Although reports show some trust in agents handling tasks like automatic IP blocking, 63% of respondents feel agents require more human oversight. Notably, nearly half of B2B companies have implemented agents without granting them actual autonomy, undermining efficiency and productivity. Experts argue that excessive caution may ultimately impede the potential savings from AI workflows, as human oversight could create bottlenecks. Furthermore, there is a call for organizations to limit agents’ roles to ensure their reliability, encouraging the use of rigorous guardrails and governance to balance efficiency and risk.
👉 Pročitaj original: CIO Magazine
