Governance in the age of agentic AI
The Governance Institute of Australia today releases Governing in the Age of Agentic AI, a white paper developed with input from Mallesons, SEEK, Diligent, EthicAI, and the University of Melbourne’s Centre for AI and Digital Ethics.
Just over half of organisations are estimated to have deployed AI agents, boosting customer outcomes while driving productivity and efficiency across the workforce.
But poorly governed agents can operate outside intended parameters, exposing organisations to cyber intrusion, financial fraud and large scale data breaches.
Governing in the Age of Agentic AI has been developed to help business leaders and governance professionals move forward with the right framework and practical tools they need to do it well.
Governance Institute has identified five key priorities for organisations to focus on:
- Clear accountability: An owner for every agent deployed.
- Meaningful oversight: Substantive human review at the systems level.
- Defined authority boundaries: Explicit limits on what agents can and cannot do.
- Transparency and traceability: Decisions that can be audited and explained.
- Capability uplift: Governance literacy that keeps pace with adoption.
Regulators warn over cyber risks
ASIC, APRA and the Australian Signals Directorate have warned that as AI becomes more autonomous, risk is moving faster and is less predictable, requiring stronger oversight and more robust cyber and resilience controls.
Agentic AI creates a new layer of legal risk, from privacy and data exposure to agents entering into contracts or breaching regulatory obligations. It will also exacerbate existing risks and legal compliance issues.
Even as liability for agentic outcomes is yet to be tested in Australia, organisations and their directors should assume they will be held accountable.
The key takeaway is to move forward, but ensure governance keeps pace so AI can be deployed with confidence and control.”Bryony Evans, Mallesons Partner
Sponsored by:
 |
 |
 |
 |
 |
 |