The Digital Transformation Agency (DTA) in Australia has recently issued interim guidance to public servants regarding appropriate AI usage. This marks the government’s initial attempt to establish guidelines for tools like ChatGPT, Google Bard, and Bing AI. The guidance aims to promote responsible and safe technology use, minimise harm, and achieve fair and reliable outcomes for all Australians.
Guidance on AI usage in government
The DTA’s guidance offers recommendations on how public service leaders should monitor and manage the use of generative AI tools within their organisations.
Public servants are advised to evaluate the potential risks and benefits for each use case before utilising publicly available generative AI platforms, such as ChatGPT, Bard AI, or Bing AI. This assessment ensures that the appropriate precautions and safeguards are in place.
The guidance provides three examples of practical use cases for generative AI in government work. These include checking the development of project plans, generating template slides for presentations, and confirming technical requirements for tender documents. For each use case, the DTA has outlined specific guidelines to follow, such as refraining from entering sensitive or classified information into public AI platforms.
Challenges and considerations
While the guidance sets out parameters for AI use, there are challenges and concerns associated with its implementation.
One of the major concerns is maintaining data security and ensuring accountability for AI-based decisions and actions. Public servants must handle large amounts of government data and classified information with care to prevent any breaches or misuse.
There are also concerns that over-reliance on AI may lead to a lack of autonomy in public service decision-making. Public servants’ understanding and familiarity with AI can also impact its effective use. Thus, it is essential to address these issues through proper training and education.
Australia’s guidance follows the global trend of providing guidelines for using AI in government settings. The UK Cabinet Office recently released its own formal guidance on generative AI, which prohibits using AI tools to draft papers about policy changes and sets limitations on producing numerical analysis.
As governments worldwide explore AI regulations, public servants must adhere to their respective agency’s policies and guidance. By doing so, they can effectively navigate the transformative potential of AI while upholding the highest ethical standards in serving the public.