Transparency key for government AI initiatives: Dominello

NSW government working on an AI user guide for agencies

artificial intelligence ai ml machine learning abstract
dny59 / kentoh / Getty Images

New South Wales customer service minister Victor Dominello says acceptance of government use of artificial intelligence (AI) will have to be built on what he describes as the “four pillars of trust”: Privacy, security, transparency and ethics.

“I think trust is always an issue,” Dominello said when questioned at an event for media and startups held last week in Sydney whether the fallout from the federal government’s ‘robodebt’ scheme would have a flow-on effect on acceptance of AI and algorithm-driven decision-making in general.

“Good customer experience has to be built on trust,” the minister said. “And that good customer experience, then builds up trust — [but] I can't give you a good customer experience unless the trust is in place first.”

The “biggest question” over the next five to 10 years as AI adoption hits its stride is “transparency,” Dominello said.

“Who is ultimately making the decision and what are the biases in?” the minister said. “Because bias in will equal bias out. And if I can see that... I can be informed of my rights, I can be informed of the outcome, and I can then choose a certain pathway. So for AI, I think the key word is transparency.”

The state government is in the throes of preparing an AI strategy. That will be accompanied by an AI ethics framework. Both documents are expected to be published in February 2020

Late last month the NSW government staged an ‘AI Thought Leaders Summit’ in Sydney. In the lead-up to the summit, the government released its draft principles on the use of AI to attendees:

  • “Community benefit - AI should deliver the best outcome for the citizen, and key insights into decision-making”
  • “Fairness - Use of AI will include safeguards to manage data bias or data quality risks”
  • “Privacy and security - AI will include the highest levels of assurance”
  • “Transparency - review mechanisms will ensure citizens can question and challenge AI-based outcomes”
  • “Accountability – decision-making remains the responsibility of organisations and individuals”.

“While there was high level agreement with the principles, and the overall policy approach, the Summit emphasised that a principles-based approach would not be possible without careful application, and without an assurance and review mechanism,” stated a communique released in the wake of the summit.

“In relation to the development of an AI User Guide for NSW Government agencies, the Summit noted that the application of principles should be clearly articulated throughout the guidance material, so that those preparing a case for the use of an AI solution can more easily make an informed decision about whether that solution will meet NSW Government standards.”

The communique said that attendees believed AI could be a “be a powerful tool for problem-solving and informing decision-making” but that “accountability must remain with agencies and government”.

“It is important that any technology that impacts our communities be subject to careful and regular human review,” states the document.

“That is, AI will provide support for government decision-making, and government must retain accountability for decisions about service delivery.”

Copyright © 2019 IDG Communications, Inc.

How to supercharge Slack with ‘action’ apps
  
Shop Tech Products at Amazon