CSIRO, Alphinity to set AI standard for ESG

CSIRO and investment manager Alphinity will partner to create a public framework to help inform ESG investment decisions about industry rollout of artificial intelligence.

The report, to be released later this year, will assess and report on responsible AI risks across multiple sectors and identify current best practice. It comes amid growing regulatory interest in AI. Picture: iStock

CSIRO and investment manager Alphinity will partner to create a public framework to help inform ESG investment decisions about industry rollout of artificial intelligence.

In a report to be released later this year, the pair will assess and report on responsible AI risks across multiple sectors and identify current best practice.

“AI will present significant opportunities to improve company performance, but we foresee potential risks in areas of governance, social licence, and operations, and investors will increasingly need to identify these,” Alphinity’s head of ESG Jessica Cairns said.

A recent study showed that corporate Australia is largely unaware of regulatory and ethical responsibilities in relation to artificial intelligence.

The announcement comes amid growing regulatory interest in AI, with the federal government having recently announced an eight-week consultation period into AI regulation.

Companies are hungry for guidance, CSIRO and Alphinity said.

“For instance, miners … aren’t necessarily developers of AI,” Ms Cairns said. “The actual risks through applications, so things like data privacy, cybersecurity, human capital impacts to their workforce, bias, social licence, all those sorts of things – that’s going to sit with the mining company.

“For investors, there are so many emerging ESG issues … it’s always important for us to get ahead of where we think companies should be going, rather than reacting after the case.”

CSIRO’s Data61 – the data and digital sub-unit of the national science agency – said AI developers wanted to understand risks in order to roll out products more quickly.

“From the technology side, because so many companies we are working with, they have AI solutions they cannot put out to the market because they don’t understand the risks,” CSIRO Data61 research director Liming Zhu said.

“I got one head of an AI lab from a major organisation telling me 90 per cent of their AI solutions cannot see the light of day, simply because they’re afraid of the risks it’s bringing and they cannot control the risks.

“Many of the AI solutions everybody’s investing in could be pushed into the market being adopted safely much faster than sitting on top of them and worrying about the risks.”