Privacy key to building trust in AI, say experts

New Zealand businesses can reap transformational benefits from AI but ensuring "privacy guardrails" are put in place is key to successful adoption – that was a message a group of AI and Privacy experts shared to an audience of board directors at the Institute of Directors Leadership Conference.
 
"Privacy builds trust, and trust enables innovation," said Deputy Privacy Commissioner Liz MacPherson, one of four panellists invited to speak on the topic of "Privacy in the Age of AI".
 
However, she believed many New Zealand businesses aren't ready to adopt AI from a privacy perspective.
 
"AI is data hungry. If you do not understand what data you hold, how it flows through systems and how it is accessed, you don't understand your risk profile."
 
"AI doesn't create new privacy risks, but it does put your current risks on steroids," she continued.
 
Fellow panellist Alastair Miller, Principal Consultant at Aura, agreed with this sentiment.
 
"Before adopting AI, you should consider whether sensitive data should be segmented away from AI. For example, PII (personal identifiable information) should be kept away from any public AI because it will be added to the training pool and could be mined by outsiders, causing a serious privacy breach."
 
Despite this, Miller said that AI is being rapidly adopted by organisations, often without any clear strategy.
 
"We're seeing AI being adopted from the bottom up, with employees using it on their own volition. It's also being incorporated into a lot of existing enterprise technology, so it is really important to have proper risk mitigation and policies in place that create "privacy guardrails."
 
Opportunity vs Risk
While privacy considerations should be front of mind for boards and directors, AI opportunities shouldn't be ignored, or New Zealand Inc. could be left behind.
 
Matt Ensor, founder of FranklyAI and Chair of the AI Forum Working Group on LLM AI, said that beyond just productivity gains, AI had the potential to introduce transformational changes to businesses.
 
"AI is introducing the ability to replace human skill and knowledge. People who are less experienced and less skilled are using AI in transformational ways – and as a result we're seeing the performance lifting in the bottom 50% of the organisation."
 
Pointing out that other countries are now adopting AI rapidly, Ensor suggested that New Zealand could be left behind if we don't leverage the opportunities of AI.
The key to benefiting from AI, agreed panellists, was making sure the organisation adequately addresses risks.
 
Microsoft Technology Strategist, Hilary Walton, said that responsible adoption of AI was key.
 
"Our customers are looking at how AI aligns to their values before adopting it."
 
Citing Microsoft's own journey to incorporating AI into technology solutions, Walton says the benefits and harms of AI should be assessed before launching anything new.
 
Over 70% of employees are already using AI, Walton says, suggesting mandatory AI training should be considered to bring people up to speed on responsible usage.
 
Action for boards
If boards aren't sure how to govern AI adoption, a good place to begin is by asking executive teams how AI is being reviewed from a strategic perspective, including consideration to the privacy risks associated with improper usage.
 
"Often, we see businesses trying to "crowbar in" AI because they think they need it, when actually it doesn't align to their strategy. How is AI actually going to benefit the business?," said Miller.
 
"If you let AI loose on all the data in your organisation (without proper strategy), it is a recipe for disaster," he continued.
 
Walton says education is key for board members, as an appreciation of AI's potential benefits and risks will help with decision-making.
 
"AI is set to transform "every room of the house" – or all departments across your organisation. As a board or executive member, you need to understand how AI might change processes, roles and responsibilities, and even your entire business model."
 
"There is AI Governance, and then there is Governing AI – how is the board enabling AI to be used in ways you haven't even considered?"
 
While all panel members agreed that AI holds significant potential and benefits for New Zealand businesses when it comes to productivity and speed to market, the emphasis was on proper adoption.
 
"To be ready for AI, you need to be privacy mature," concluded MacPherson.