EDB Pushes Energy Efficiency as Core Requirement for Agentic AI

At the United Nations Global Compact Leaders Summit, EDB Chief Legal Officer Rob Feldman argued that the industry must move beyond treating sustainability as an afterthought. New research from the company reveals that 69% of enterprise decision-makers now weigh energy efficiency as heavily as cost and reliability when selecting AI platforms. To address this, EDB is advocating for a sovereign AI operating system that embeds governance directly into the data layer, treating efficiency as a fundamental design principle rather than a post-deployment optimization.

Designing for Efficiency

The company is promoting an approach dubbed "intelligence per watt," which aims to reduce unnecessary compute and data movement. Benchmarks using EDB Postgres AI demonstrated that consolidating workloads could reduce token consumption by 57% in specific deployments. Independent analysis by Incendium Consulting further suggested that such architectural shifts could lead to 94% fewer compute cores and 87% lower emissions for large-scale operations. Feldman noted that these gains prove performance and environmental responsibility do not have to be a trade-off, urging organizations to adopt transparent reporting frameworks to ensure that autonomous technology remains accountable as it scales toward an expected one billion agents by the end of the decade.