AI can help charities, but proper oversight is essential

By CathNews New Zealand

September 5th, 2025

(Jernej Furman/Wikimedia Commons)

AI can help New Zealand charities work more efficiently and effectively, but an expert warns the trust, privacy and wellbeing of those served must be safeguarded. Source: The Post.

In a recent opinion article in The Post, Cloud specialist Jo Murphy noted that, in charities, the adoption of AI often happens organically. Volunteers or staff members can experiment with the new technology.

“But without shared guidelines or a common understanding, there’s a chance for unintentional missteps, particularly when it comes to handling sensitive data, protecting privacy or avoiding bias,” she wrote.

Ms Murphy said she had “recently heard of a volunteer who used ChatGPT to draft client case notes without knowing the information could be stored or processed outside the organisation’s control. Intended as a time-saving measure, it highlights how everyday, well-intentioned use of AI can still create risks if there are no clear guidelines or safeguards in place.

“When AI is used without proper oversight, charities face several significant risks. One major concern is the potential for privacy breaches, particularly when sensitive data is fed into public AI systems. There is also a risk of output bias, especially when AI is used to generate narratives about individuals or communities, which can lead to harmful or inaccurate portrayals. Additionally, if funders or service users become aware that AI is being used inappropriately, it can result in a serious loss of trust.

“By setting clear expectations, offering training and providing safe spaces for staff and volunteers to explore AI, boards can help ensure it becomes a tool that supports their overall goals,” Ms Murphy added.

“Strong governance is essential. Without a clear, organisation-wide picture of how AI is being used, boards cannot properly assess risks, build staff capability, or put effective guardrails in place. Increasingly, forward-thinking charities are treating AI as they would any other critical system – defining acceptable use, setting boundaries and assigning responsibility for oversight. It all begins with one deceptively simple step: asking the right questions.”

FULL STORY

Charities using AI: beware efficiencies don’t come at the expense of privacy and integrity (By Jo Murphy/The Post – subscription required)