8AAdditional expectations—provider will take reasonable steps regarding generative artificial intelligence capabilities

8A Additional expectations—provider will take reasonable steps regarding generative artificial intelligence capabilities

(1) If the service uses or enables the use of generative artificial intelligence capabilities, the provider of the service will take reasonable steps to consider end‑user safety and incorporate safety measures in the design, implementation and maintenance of generative artificial intelligence capabilities on the service.

(2) If the service uses or enables the use of generative artificial intelligence capabilities, the provider of the service will take reasonable steps to proactively minimise the extent to which generative artificial intelligence capabilities may be used to produce material or facilitate activity that is unlawful or harmful.

Examples of reasonable steps that could be taken

(3) Without limiting subsection (1) or (2), reasonable steps for the purposes of this section could include the following:

(a) ensuring that assessments of safety risks and impacts are undertaken, identified risks are appropriately mitigated, and safety review processes are implemented throughout the design, development, deployment and post‑deployment stages of generative artificial intelligence capabilities;

(b) providing educational or explanatory tools (including when new features are integrated) to end‑users that promote understanding of generative artificial intelligence capabilities on the service and any risks associated with the capabilities;

(c) ensuring, to the extent reasonably practicable, that training material for generative artificial intelligence capabilities and models do not contain unlawful or harmful material;

(d) ensuring, to the extent reasonably practicable, that generative artificial intelligence capabilities can detect and prevent the execution of prompts that generate unlawful or harmful material.