Pearcey Institute submission to Australia's AI Ethics Consultation

31 May 2019 Canberra

DIIS AI Consulation

The Pearcey Institute submitted feedback to the Department of Industry, Innovation and Science public consultation on Artificial Intelligence: Australia's Ethics Framework discussion paper released in April 2019.

Pearcey members with expertise in the developing international approach on ethical frameworks for AI, attended the briefing sessions and and reviewed the paper. Our feedback includes responses to the questions asked in the discussion document, as well as broader views on the challenges of developing this framework based upon our collective experience of the evolution and adoption of technology in all aspects of our lives to-date.

In summary, the following high level observations were made:

  1. A concern about the challenges around defining "Artificial Intelligence", and
    furthermore defining the workforce that would be impacted by an AI framework.
  2. A recommendation to consider a layered solution, cascading from high-level
    considerations to the minutiae of specific use-cases, similar to a number of Risk
    Assessment frameworks.
  3. In addition to the introduction of machine-driven decision making, we raise the
    related topic of unauthorised reuse of data acquired without our knowledge.
  4. A concern about the pragmatics of applying an AI ethics framework, and how it
    will be used ("enforced").
  5. A recommendation to align the framework for AI to existing frameworks.