The Australian Taxation Office (ATO) is seeking a sophisticated artificial intelligence system to detect and prevent fraud across its digital services in real-time, as part of a major upgrade to its security infrastructure.
The system aims to protect more than 14 million users who access ATO online services, with usage reaching peaks of 230,000 users per hour during tax time.
ATO said the procurement forms part of its Counter Fraud Program, which is implementing new capabilities to combat identity-enabled fraud attacks on its systems.
"The system should be available all the time to ensure no transactions are processed without proper risk analysis being undertaken," the tender documents state.
The selected solution must operate 24/7 and handle up to 800 interactions per second while meeting strict government security requirements for handling protected data.
Vendors must demonstrate experience in fraud prevention research and development, maintain a minimum system uptime of 99.99 per cent, and ensure all staff have baseline security clearances.
The system must comply with the Australian Signals Directorate's Essential Eight security controls at Maturity Level 2 or higher and undergo an Information Security Registered Assessors Program assessment.
Technical requirements include real-time API integration with ATO digital platforms, support for machine learning models, and ability to process both structured and unstructured data.
The contract will run for an initial 12 months with five one-year extension options. The ATO expects the system to be tested by December 2025 and operational by April 2026.
All data must be hosted within Australia to comply with sovereignty requirements.
Interested vendors can submit proposals until March 15, 2025.
-1.jpg&w=100&c=1&s=0)
.png&w=100&c=1&s=0)
