Superannuation funds must address several key risks as they digitally transform and implement artificial intelligence (AI) solutions and governance frameworks.
Speaking at an international business review webinar, Professor Matthew Kuperholz, chief data scientist, PwC, said technology like artificial intelligence was neither good, bad or neutral.
“It’s what we do with it that counts and we need to acknowledge that the accountabilities of being responsible with this technology… is squarely on us – the decision makers,” Kuperholz said.
He said super funds should start by assessing the performance risk of the AI by looking at whether it was stable, accurate, biased, morally, and legally acceptable and if it could be explained to members.
He said economic risks would also need to be considered.
“We're displacing people. We're changing the profile of the liabilities,” he said.
Trustees would need to account for security risks as well, according to Kuperholz, including the risk that the AI could be reverse engineered, used for nefarious reasons or have inadequate data security.
“Can we even detect that the AI has drifted and gone rogue?” Kuperholz said.
“If we change the shape of our workforce, how do we get people back in the loop when it's gone awry?”