Organizations must overhaul their infrastructure in order to maximize the agentic AI opportunity (www.techradar.com)

A recent survey by Google reveals that 83% of IT leaders believe their organizations need significant infrastructure upgrades to effectively deploy agentic AI. With over 1,400 senior IT leaders polled, the report highlights the challenges of operating on legacy systems, leading to unexpected costs linked to increased power consumption and operational complexity. Key barriers include security concerns, governance issues, and a rise in hidden operational costs as organizations attempt to scale AI on outdated infrastructure. To address these challenges, Google suggests several initiatives, including leveraging fluid compute to optimize resource allocation, using enterprise-grade governance tools like Agent Gateway for improved oversight, and adopting a unified data layer to streamline data access and reduce storage costs. Additionally, the shift toward edge computing is recommended, with 90% of organizations considering this strategy to achieve reduced latency and better resilience. As energy cost becomes a critical factor, 91% of leaders are prioritizing power consumption in their hardware selections, marking a pivotal moment for organizations to reconsider their approach to AI infrastructure.