Ms Fiona McLeod SC presented the 2019 Solomon Lecture at the Queensland State Library, as part of the Office of the Information Commissioner’s celebrations marking 10 years of the Right to Information Act 2009 and Information Privacy Act 2009 in Queensland.
The lecture focused on privacy and accountability in the context of increasing adoption of technologies such as artificial intelligence (AI) in government functions and decision-making.
Fiona detailed a number of challenges facing accountability in the world of AI.
AI is a black box. We don’t know how it works.
Absence of soul. Machines are inherently unethical. Bias is developed due to lack of holistic data, and programs, like programmers, are not free from bias.
All rights are forfeit. Terms and conditions too complex. Marketing and third-party use of personal data is often inherent in terms and conditions. De-identification is inadequate and often users have no choice but to give personal details for services, particularly those provided by government.
Nothing is sacred. Everything is hackable and nothing is safe. Data pools containing personal information are particularly vulnerable.
The government gets what the government wants. National security claims currently take precedent over all other, including human rights and freedom of press.
Protection laws are not fit for purpose. It has been too long since the Australian Law Reform Commission has review of privacy legislation.
At the close of the lecture Fiona identified ways to hold back the tide so we can protect privacy and embed accountability including:
greater education and digital literacy
a safety-first approach
stronger integrity frameworks
implement a robust integrity committee
A panel that included the likes of Scott McDougall (Human Rights Commissioner), Dr Brent Richards (Medical Director of Innovation, Gold Coast Hospital and Health Service), Mr Simon McKee (Deputy Commissioner, Queensland Treasury), Phil Green (Privacy Commissioner) and Madonna King (Master of Ceremonies), followed Fiona’s lecture providing additional insights into how to protect privacy and accountability in the world of AI including:
provide low tech and no tech alternatives
create a certification of human rights for AI products
If you are looking for further information on this topic see the resources below: