Queensland Implements Obligatory Risk Assessments for AI and Automated Decision-Making
We independently review everything we recommend. When you buy through our links, we may earn a commission which is paid directly to our Australia-based writers, editors, and support staff. Thank you for your support!
Queensland Requires Risk Assessments for AI and Automated Decision-Making Systems
Quick Read
- Queensland will require risk assessments for AI and automated decision-making (ADM) in public sector initiatives.
- Evaluations conducted internally and examinations by external parties will be necessary.
- The AI governance policy and its accompanying AI risk assessment framework are in the final stages of development.
- Possible implementation of the ISO42001 AI Management System Standard.
- Current frameworks, such as the ICT Investment Review, will be incorporated into the new policy.
- QGCDG currently assists departments with the planning and implementation of AI projects.
- Examples of AI initiatives include employing drones in farming and the QChat conversational agent.
Implementation of Compulsory Risk Assessments
The Queensland government plans to implement compulsory internal assessments and external evaluations for public sector projects that utilize artificial intelligence (AI) and automated decision-making (ADM). The objective of this initiative is to identify and manage the unique risks linked to these technologies.
Formation of AI Regulation Strategy
As stated by Chris McLaren, who holds the position of Queensland’s Chief Customer and Digital Officer, the Queensland Government Customer and Digital Group (QGCDG) is nearing completion of an AI governance policy. This policy is accompanied by an AI risk assessment framework that aims to direct the implementation of AI within multiple government sectors.
Adopting Industry Standards
The QGCDG is also considering the implementation of the ISO42001 AI Management System Standard along with other relevant industry standards. This would facilitate the identification and management of AI risks throughout their lifecycle within the Queensland government.
National Framework Agreement
All the ministers responsible for data and digital matters from the federal and state territories recently concurred on a national framework that advises, but does not require, assurance scoring and controls for government projects. This framework highlights the importance of internal evaluations and external reviews, dependent on financial and risk criteria.
Existing Implementations in Additional States
New South Wales (NSW) and Western Australia (WA) are presently the sole states that have implemented these types of systems. NSW’s system pertains exclusively to projects incorporating AI, while WA’s encompasses both AI and ADM projects.
Range of the New Framework
Although QGCDG’s framework is currently in its finalization stage, it is set to encompass both AI and ADM in extensive projects. This new policy will be incorporated into existing mandatory systems like the ICT Investment Review and the portfolio, program, and project assurance framework.
Continuous Assistance for AI Initiatives
Even though new frameworks have been introduced, QGCDG has been assisting departments in the planning and implementation of AI projects. The existing assurance process considers risk factors associated with AI and ADM. When needed, risks are addressed with the Digital Economy Leaders Sub-Group, responsible for overseeing digital investments.
Examples of Ongoing AI Initiatives
One significant initiative features the Department of Agriculture and Fisheries utilizing drones with classification algorithms to detect weeds in sugarcane fields. Another instance is QChat, a supportive chatbot implemented in multiple departments to enhance customer facilitation, regulatory technology, productivity, and cybersecurity.
Future Assurance Measures
All artificial intelligence initiatives under the Queensland government are presently documented with the QGCDG. They use an assurance profiling instrument to decide if just an internal evaluation is sufficient or if an extra external review is needed. The upcoming policy will establish the assurance check requirements for forthcoming projects.
Summary
The Queensland government is on the verge of introducing compulsory risk assessments for AI and ADM in public sector initiatives. The goal of this measure is to guarantee that risks are properly assessed and managed. Supported continuously by QGCDG, various departments have started incorporating AI into diverse applications, such as agriculture and customer service. The upcoming policy will offer an organized framework for future technology implementations, aligning with national standards and improving oversight.
What is the primary goal of Queensland’s newly implemented policy regarding AI and ADM?
The primary goal is to assess and reduce risks linked to AI and ADM in public sector initiatives by requiring both internal evaluations and external audits.
What criteria could Queensland establish for governing AI-related risks?
Queensland is considering implementing the ISO42001 AI Management System Standard in addition to other pertinent industry standards.
Which states have already adopted comparable systems?
New South Wales (NSW) and Western Australia (WA) have implemented comparable systems. NSW concentrates exclusively on AI initiatives, whereas WA encompasses both AI and ADM initiatives.
How will the new policy be incorporated into the current frameworks?
The new AI risk assessment framework is set to be integrated into current mandatory systems such as the ICT Investment Review and the portfolio, program, and project assurance framework.
Could you give some examples of ongoing AI projects in Queensland?
Illustrations encompass the Department of Agriculture and Fisheries utilizing drones to identify weeds in sugarcane plantations and the QChat chatbot employed across multiple customer service scenarios.
What current support does QGCDG offer for AI projects?
QGCDG assists departments by providing expert consultants in data and artificial intelligence to assess and minimize risks during the planning and implementation of AI projects.
What will the upcoming policy outline concerning assurance checks moving forward?
The upcoming policy will detail the degree of assurance checks required for governmental AI and ADM initiatives moving forward.
