Back to Articles
Sportsbet Deploys AI Gateway to Enforce Guardrails and Manage Costs

iTnews

ENRICHED

Details

Date Published
22 May 2025
Priority Score
2
Australian
Yes
Created
7 Aug 2026, 04:00 am

Authors (1)

Description

Starts with existing agents and coding assistant traffic.

Summary

Sportsbet has implemented a centralized AI gateway via TrueFoundry to manage large language model traffic and agent tool calls. The technical architecture focuses on real-time validation guardrails and policy enforcement intended to maintain oversight of AI workloads without increasing latency. While this represents a significant enterprise-level deployment of AI safety tools in the private sector, the primary focus is on operational governance, cost attribution, and corporate compliance rather than mitigating catastrophic or existential AI risks. The deployment highlights how Australian commercial entities are operationalizing safety guardrails for internal productivity tools and coding assistants.

Body

Sportsbet has deployed a centralised control plane to govern its increasing adoption of AI and control the technology’s usage and costs. The company has deployed the TrueFoundry AI gateway on its AWS infrastructure, and is already running some “immediate” use cases through the gateway. The gateway is designed to act as a “centralised control plane for all LLM traffic, MCP-connected agent tool calls, and AI workloads across Sportsbet's engineering teams,” the vendor said in a statement. It also runs “validation guardrails in parallel with [AI] model calls, so policy enforcement adds no measurable latency to the critical path.” So far, Sportsbet has replatformed an existing AI agent to run via TrueFoundry “ to gain usage metrics, guardrails, and MCP-routed tool visibility,” the two companies said. The agent is “used by business analysts and product managers to draft product requirement documents and user stories via Jira integration.” Coding assistant traffic is also now being routed through the gateway. With centralised governance, Sportsbet is looking essentially at enforcing guardrails and manage costs via “full visibility into model usage, token consumption, and cost attribution per application and team.” "AI is reshaping how we build products and customer experience and Sportsbet intends to lead, not follow,” Sportsbet’s general manager of data and AI Niall Keating said. “The hard part was never the pilot; it's turning AI into a core capability that runs the business at scale.”