The NZ data and tech head believes that good governance should enable rather than hinder AI adoption
In light of struggles to manage AI value chains, proper governance has become critical to maintaining operations.
“The research reflects the underlying challenge: many organisations are already relying on suppliers they cannot properly oversee or audit. That does not mean governance stops at the contract. In fact, it makes governance more important”, Nick Valentine, DLA Piper’s New Zealand head of data and technology, told NZ Lawyer.
Observability is a key trait of strong AI governance, he pointed out.
“Can the organisation show who owns the system, what risk rating applies, what data and suppliers it depends on, how outputs are tested, what human oversight exists, and what happens when something fails?” Valentine asked.
He explained to NZ Lawyer that governance “should enable AI adoption, not block it”. However, “if there is no inventory, no clear owner, no testing record, no supplier due diligence and no escalation pathway, that may indicate the organisation’s governance aspirations have not yet translated into operational controls”.
Valentine urged organisations to “stop treating AI as an IT project”.
“AI is becoming a new layer of business infrastructure, so accountability needs to sit where enterprise risk sits”, he told NZ Lawyer. “Where AI creates material enterprise risk, the board should have appropriate oversight and ask the hard questions. Management should appoint a senior accountable owner or AI governance forum with real decision rights. Then legal, privacy/cyber, procurement, HR, risk and commercial stakeholders should work through a common approval and monitoring process”.
The goal is to establish clear accountability, Valentine indicated. According to DLA Piper, 51% of Australia and New Zealand respondents to its The AI Year study believed that skilled or specialist oversight could improve the effectiveness of their organisations’ AI governance.
“The point is not to centralise all AI expertise in one team. It is to have one clear point of accountability, supported by a distributed operating model. AI is too strategic and too interconnected to be left in a silo”, Valentine told NZ Lawyer.
In spite of risk management worries, Australia and New Zealand organisations are still positive about AI, with 75% of respondents recording overall confidence in performance stability, investment returns, valuations, and AI’s ability to lead to expected business gains.
