Geordie launches Cost Intelligence to tie AI spending to agent activity

Geordie AI Ltd., an agent security startup, has introduced Cost Intelligence, a new feature that links AI spending to the agents and workflows that generate it. This capability extends beyond security, focusing on financial governance. Traditionally, AI spending is measured by consumption, showing which models ran and the number of tokens used, but it doesn't identify the agent responsible for the charge.

However, Geordie has quantified the issue. For example, one retailer was losing $1 million a month using more expensive models for tasks that did not require them, while a bank saw an agent consume 50% of its monthly AI budget in a single day due to a tool loop. Geordie's platform already closely monitors agents for security and governance, providing insights into agent behavior, token usage, and the models called.

This new platform can now trace cost changes back to the specific agent, task, and owner. The release includes five key capabilities: aggregating spending from all connected platforms into a single view and attributing it to the corresponding work; allowing teams to drill down from an organization-wide total to more detailed levels such as platform, team, user, model, individual agent, or workflow; tracking caching efficiency per agent to identify under-caching; mapping token efficiency by comparing spend against volume, ensuring busy agents are not unfairly penalized; and detecting anomalies like infinite loops, oversized model usage, and inefficient sessions.

According to Henry Comfort, Geordie's co-founder and CEO, "Tokens and invoices tell you what you consumed; they do not tell you what work drove that consumption." With Cost Intelligence, organizations now have the foundation to understand agent ROI. Jacob Sweat, vice president of technology at Wood Partners LLC, highlighted the importance of this visibility, stating that once they could see the cost of their agents and the value they provided, their internal focus shifted from how much was being spent to whether the spend was worthwhile.

Hanah-Marie Darley, Geordie's chief AI officer, emphasized that the risk from autonomous agents extends beyond security, stating that the risk from autonomous agents "isn’t limited to security alone." Geordie AI, a venture capital-backed startup, has raised $36.5 million since its September 2025 launch, with $30 million from its Series A round led by Balderton Capital and $6.5 million from its seed round led by Ten Eleven Ventures and General Catalyst Group Management.