Ico Maker - stock.adobe.com

Australian firms coming to grips with AI ethics

Australian businesses are warming up to AI, but just two in five have standards and guidelines for AI ethics, a study finds

Most Australian businesses are already tapping artificial intelligence (AI) to improve key business outcomes, but just two in five have standards and guidelines in place to mitigate the negative impact of the technology, a survey has found.

According to the survey conducted by Fifth Quadrant on behalf of LivePerson, which specialises in asynchronous communication, AI and automation, two-thirds of Aussie firms are now leveraging AI to drive positive outcomes for their organisations, employees and customers.

In fact, the 562 Australian businesses that participated in the study expect a 12% growth in revenue, or an average increase in revenue of more than A$1.8m over the next 12 months after achieving broad AI usage.

Further, a significant number of Australian businesses broadly using AI said the technology has had a positive impact on customer satisfaction (65%) and customer retention (61%), employee satisfaction (59%) and employee productivity (69%).

The enthusiasm around AI, however, has raised new ethical concerns, including the potential for AI technologies to fall into the wrong hands (85%), loss of privacy (85%) and unauthorised access to data (84%).

In April 2019, the Australian government released a public consultation paper on AI ethics developed by the Commonwealth Scientific and Industrial Research Organisation (CSIRO) that spells out the core principles of AI.

More than 90% of Australian businesses generally viewed these principles as “somewhat” or “extremely” important, including the principles of privacy protection, regulatory and legal compliance and that AI must not be designed to harm or deceive people.

Meanwhile, in the absence of a formal framework, Australian companies are taking steps to mitigate the risk of negative outcomes from the use of AI.

Some 36% said they were doing risk assessments, while 31% were monitoring industry standards and conducting ethics training for employees, and 30% were developing best practices and conducting impact assessments at the same time.

Australian firms had mixed views when it came to accountability of AI deployments, with an equal number of survey respondents (46%) pinning accountability on those developing and deploying the technology.

Within their organisations, respondents noted that company leadership, including C-suites (34%) and boards of directors (34%), were most likely to have ultimate accountability for the decisions made by AI systems.

From a regulatory perspective, two in five businesses believed an Australian government body should be responsible for setting and enforcing AI ethics. However, only half of them thought it was “extremely important” for AI systems to comply with international and Australian regulations.

“We’re on the cusp of a new era,” said Rob LoCascio, CEO and founder of LivePerson. “AI will transform people’s lives by making technology more efficient, easier to use, reliable and capable, opening up tremendous human potential.

“Progressive companies are already starting to achieve this. However, business leaders should approach AI with their eyes wide open, looking at practical and proactive measures to ensure ongoing ethical implementation that results in the best outcomes for customers.”

In autonomous driving, for example, there is a need for more transparency around the algorithms that power autonomous vehicles, said Kate Marshall, a partner in KPMG Law in Australia focusing on privacy, technology and data.

“There’s a need to understand the ethics and bias behind them, how they work and whether they are in line with community expectations,” she added. “This is where KPMG is getting involved – we’re hosting an AI and ethics forum, bringing together industry leaders, academics and regulators to talk about that. I think it’s a really important conversation.”

According to KPMG’s Autonomous vehicles readiness index 2019, Australia ranked 15th globally in terms of the level of preparedness for autonomous vehicles.

It gained top marks in the “supportive regulations” category – amid broader signs of positive policy and legislative changes – and scored highly on connectivity infrastructure.

“Australia has performed well in certain areas but still needs improvement on others,” said Praveen Thakur, partner for management consulting at KPMG Australia.

“For example, our lower rating on technology and innovation will hopefully be boosted in future by the recent establishment of a federal body for transport technologies, which aims to unify the states’ and territories’ governments and agencies in delivering future transport technologies.

“As a federation it is crucial that different levels of government work together effectively and we are seeing this via our strong rating in the survey’s policy and legislation category.”

Read more about AI in Australia

  • Australia is being transformed by artificial intelligence and digitisation, particularly in areas such as education and public safety.
  • International teams at a robotics competition in Australia have developed robots that prevent spent nuclear fuel from falling into the wrong hands.
  • Australia does not need a robot tax, but it needs to prepare for a rising tide of automation, according to the head of the country’s top technology research body.
  • As the enthusiasm for AI gathers pace, Australia’s chief scientist sounds out a note of caution and calls for more regulation of AI.

Read more on IT legislation and regulation