Cause and effect: AI use in the charity sector
Artificial intelligence (AI) is no longer a future consideration for charities. It is already part of daily operations across the sector. Infoxchange’s 2025 Digital Technology in the Not-for-Profit Sector Report (Infoxchange Report) found two-thirds of not-for-profit organisations now use generative AI, principally for content creation, reporting and editing. Use of AI for grant writing doubled in just 12 months. At the same time, the report found only 14% of organisations have an AI policy in place – meaning adoption is significantly outpacing governance.
This disparity presents both opportunity and risk. AI can help charities stretch limited resources further, but it also raises questions of bias, privacy, transparency and public trust, which are especially acute in a sector serving vulnerable communities.
For charities, the question is not simply whether AI can deliver efficiencies, but how it can be adopted in a way that remains consistent with charitable purpose, human judgement and public trust.
The opportunity
Charities face well-documented structural pressures: rising demand for services, constrained funding, workforce fatigue and growing expectations of efficiency and accountability. AI offers a practical response to some of these pressures. The Australian Charities and Not-for-profits Commission (ACNC) guidance on charities and artificial intelligence identifies applications ranging from personalising donor communications and matching volunteers with opportunities, through to supporting disaster response and measuring program impact.
Perhaps the most immediate benefit for many organisations is automation of routine and administrative tasks. AI tools can assist with drafting and editing communications, summarising meeting notes, generating reports and troubleshooting technical issues. The Infoxchange Report illustrates the breadth of this uptake: three out of four organisations using generative AI apply it to content creation, reporting, writing and editing; almost half use it for meeting notetaking and summarising; and over 40% use it for technical assistance and troubleshooting. By reducing the time staff and volunteers spend on repetitive administrative work, these tools can free capacity for higher-value activities such as direct engagement with beneficiaries, service delivery, strategic planning and relationship-building.
Infoxchange’s report confirms 80% of organisations that have implemented AI report productivity and impact improvements.
A paper authored by Kim Weinert and Brydon Wang at the University of Queensland titled ‘Artificial Intelligence and Australian Charities: Governance, Bias, Human Rights and Public Trust for Non-Profits in the Algorithmic Age’ (UQ Paper) also observes that AI is increasingly being used in service delivery – for triage, prioritisation and needs assessment. When deployed well, these tools can support more consistent and evidence-based decisions about how limited resources reach beneficiaries.
But charities should assess these opportunities against purpose, not adopt AI simply because the technology is available or increasingly common.
From tool to governance decision
A threshold question for any charity considering AI is whether that decision is being treated as an operational matter or recognised for what it is: a governance decision. The UQ Paper frames this succinctly: the question is not whether charities can use AI but whether they are justified in doing so while remaining trustworthy.
This framing matters because AI decisions by charities are strategic. The ACNC’s guidance emphasises the purpose of AI should be to enhance human decision-making rather than replace it, and Responsible People must exercise due diligence to ensure AI initiatives are overseen by individuals with the necessary skills and knowledge.
Governance Standard 5 requires charities to ensure their Responsible People act with reasonable care and diligence, honestly and in the charity’s best interests, without misusing their position or information and with proper disclosure of conflicts. Although Governance Standard 5 does not specifically mention AI, the adoption of AI (particularly where it affects service delivery, data handling or resource allocation) clearly engages these duties. Responsible People can meet this standard by approaching AI adoption with a clear understanding of its implications, appropriate safeguards and alignment to the charity’s purpose.
Key risks for the sector
The risks of AI in the for-purpose sector (and in particular charities) differ in some ways from those in a commercial context. Charities’ beneficiaries often experience vulnerability, may have limited capacity to challenge automated decisions and may have few or no alternative sources of support. This means that errors, biases or opacity in AI-driven processes can cause disproportionate harm.
- Bias and discrimination. AI systems trained on incomplete or historically skewed data can produce discriminatory outcomes – a risk the ACNC identifies and Philanthropy Australia underscores by noting AI tends to reflect dominant narratives. Where AI participates in decisions about access to services, unchecked bias can directly harm the beneficiaries whom charities exist to support.
- Privacy, data security and cybersecurity. The Infoxchange Report found half of all not-for-profit organisations identify data security, ethics or privacy as their primary barrier to AI adoption. These concerns are well-founded, as only 23% have a documented cybersecurity plan and 14% experienced a breach in the past year.
- Human connection and digital exclusion. Direct support, advocacy and trusted human relationships remain at the heart of social services. Over-reliance on AI risks diminishing the empathy and human connection which distinguishes charitable service delivery. There is also a real risk beneficiaries without digital literacy or access will be left behind by AI-driven processes.
- Public trust. Public trust is one of a charity’s most valuable assets and should be front of mind when considering any potentially trust-damaging action. Philanthropy Australia describes transparency about AI use as non-negotiable: stakeholders and communities should know when and how AI is being used, particularly in consequential contexts such as grant scoring, service allocation or impact measurement.
Practical steps for leaders
While Australia does not yet have a formal AI regulatory framework, existing laws and governance standards still apply. Commonly recommended practical steps include:
- Map current use and set clear guardrails. Start by identifying where AI is already being used across your organisation. From there, boards and leadership teams can assess which uses carry higher risk, decide what is permitted or prohibited and put in place clear guardrails. This turns AI from an ad hoc efficiency tool into something that is deliberately governed and aligned with the organisation’s purpose.
- Keep the human connection. Work in the sector depends on trust, empathy and personal relationships. Ensure AI supports rather than replaces these qualities, both in service delivery and in the oversight of AI itself. Establish clear protocols for human review of consequential decisions and keep space for the discretion and humanity that defines the sector.
- Connect AI oversight to existing governance frameworks. Integrate AI risk assessment into your existing risk register and board reporting. Ensure someone with appropriate expertise is accountable for AI governance and regularly review AI use as part of your standard governance considerations.
Key takeaways
AI presents real opportunities for charities, but adoption should be approached as a governance decision rather than a purely operational one. The charitable sector occupies a unique space, built on trust, human connection and service to beneficiaries. This context calls for thoughtful and meaningful engagement with AI, recognising both its potential and its risks. Boards and leaders of charities are well placed to lead this conversation within their organisations, drawing on their existing duties under the ACNC Governance Standards.
Done well, AI adoption can enhance a charity’s capacity to serve its purpose. Done poorly, it risks undermining the very trust and human connection on which charitable work depends. The key lies in approaching AI deliberately, with clear governance, appropriate safeguards and a genuine commitment to the values that define the sector.
How can we help?
If you would like to discuss any of the issues raised in this article, please get in touch with our specialist Charities and Social Sector team.