Breaking free from pilot purgatory. The strategies needed to scale agentic AI (www.techradar.com)

At the recent World Economic Forum in Davos, discussions around AI have shifted from lofty ambitions to practical strategies for implementing agentic AI securely across organizations. Leaders are increasingly frustrated with isolated pilot projects and are now focused on scaling these systems into core business functions. A report indicates that nearly 75% of organizations plan to boost investments in agentic AI. However, simply increasing budgets may not lead to significant transformation without addressing key barriers—primarily security and data privacy concerns—which 50% of organizations in EMEA cite as major obstacles. The article emphasizes that trust is essential for the successful deployment of agentic AI. Nearly 70% of agentic AI decisions require human verification, highlighting the importance of human oversight while simultaneously allowing AI to execute tasks efficiently. A balance between human judgment and AI capabilities is necessary as organizations move toward integrating these systems into customer-facing roles, where risks are heightened. To transition out of pilot purgatory, organizations must build robust observability and governance structures that ensure AI operates securely and effectively, thus enabling sustained business value and innovation without sacrificing accountability.