THE City of Kalamunda is moving to regulate how artificial intelligence is used by staff, with a draft policy setting rules around privacy, oversight, and accountability.

The policy outlines how AI tools, including generative platforms such as ChatGPT, can be used within the organisation, while prohibiting the use of sensitive or confidential information.

Under the draft, AI must not be used to make decisions or deliver services without human oversight, and all outputs must be verified by a suitably qualified or experienced person before being used in decision-making or public communication.

The policy also requires staff to disclose when AI has influenced decisions, communications or services and establishes mechanisms for reviewing and challenging AI-generated outcomes.

Mayor Margaret Thomas said the move was driven by the increasing use of artificial intelligence tools, which “make it essential to establish a strong governance framework to guide its responsible use.”

“The policy ensures any current or future use of AI is ethical, lawful and in the best interests of our community.”

The draft framework also places restrictions on what platforms can be used, requiring approval from the city’s Information and Communications Technology (ICT) services before any public AI tools are adopted.

Unapproved platforms may be blocked, and all commercial AI tools must undergo security assessments to prevent uncontrolled third-party data sharing.

The policy defines sensitive information as data that could compromise privacy or security if disclosed, and explicitly prohibits entering such information into AI systems.

It also includes compliance and monitoring measures, allowing the city to verify usage through log reviews, browser history checks, and internal and external audits, with breaches potentially resulting in disciplinary action.

Mayor Thomas said the policy aimed to balance innovation with community trust.

“It sets out clear governance, accountability and risk-management arrangements so that AI can support innovation while protecting community trust, privacy and the city’s reputation,” she said.

If adopted, the policy will apply to council members, employees, contractors and volunteers using AI technologies in the course of their duties.

The city said the framework would ensure AI is used to support, rather than replace, human decision-making, while improving service delivery for residents.