Charity AI adoption: An overview of AI governance, opportunities and risks for charity trustees
13 Oct 2026 • Audit and Assurance • Charities and Not-For-Profits
Artificial intelligence is now used by almost eight in ten UK charities, according to the 2026 Charity Digital Skills Report. From administrative tasks and fundraising to the implementation of finance systems and CRMs that incorporate AI features, adoption is accelerating.
This insight looks at common use cases, hallucination, data protection, and fraud risks charities are seeing and how charities can adopt AI responsibly with confidence.
Artificial intelligence has moved from experimental interest to everyday use across the charity sector. According to the 2026 Charity Digital Skills Report, almost eight in ten charities (79%) are now using AI tools in some way. The proportion of charities actively or strategically embedding AI in their day-to-day work has also grown quickly, up to 38% from 25% the previous year. This shift reflects a move away from tentative experimentation and towards AI becoming a more established part of how many charities operate, plan and deliver their services.
Common uses across the sector
The most established use cases are administrative. Three in five charities (60%) now use AI for everyday tasks such as summarising meeting notes and drafting emails, and AI is increasingly used to create documents, reports and presentations.
AI is also playing a growing role in fundraising and bid writing. Around 45% of charities are using it to support grant applications, rising to 52% among small charities, particularly to meet strict word counts, structure proposals and get started on a blank page. Charities using it this way also note that AI-generated bids often need real rework to get the tone, message and accuracy right before they are submitted, so it is best used to support the drafting process rather than replace it.
For trustees and finance teams, one of the more notable developments is the way AI capability is beginning to sit inside the finance systems charities already use, rather than as a separate tool. AI features are increasingly being built directly into accounting and finance software to support tasks such as bank reconciliation, transaction categorisation and answering finance queries. While AI remains one of the least adopted areas of financial management, with only 8% of charities currently using it in this way, the direction of travel is clear. As AI becomes a standard feature of finance systems rather than an optional add-on, we expect adoption to increase over time.
A fast-moving landscape
The pace of change is significant too. Use of AI agents, tools that can act with a degree of autonomy, has grown from 11% of charities exploring them last year, to 28% this year. Charities that engage early with these tools — understanding how they work, their limits, and the governance they require — will be better placed to use them effectively.
Managing the risks
Wider adoption brings an increased number of risks for trustees and finance teams to consider:
Data protection remains a significant concern for many charities. Data privacy, GDPR, and safeguarding concerns are cited by 43% of charities as a barrier to AI adoption, particularly among those working with vulnerable people and children, and some have avoided using AI in client-facing work as a result. Many mainstream providers now offer enterprise-tier products that commit to not using customer data for training, though this does not remove a charity's obligations under the UK General Data Protection Regulation and the Data Protection Act 2018.
Fraud is also a significant concern. The latest Charity Fraud Report found that more than a third of UK charities (34%) experienced fraud or attempted fraud in the previous year. The report also highlighted growing concern about cyber-enabled and AI-assisted fraud, including AI-enabled impersonation, deepfakes, and phishing attacks that are becoming harder to identify using traditional warning signs. As generative AI makes it easier to produce convincing communications, supporting documentation, and digital identities at scale, charities may need to adapt their controls, review processes and staff training to ensure fraudulent activity is identified and challenged effectively.
AI tools can also produce inaccurate, incomplete, or fabricated information, often referred to as "hallucinations". While these errors can sometimes be obvious, they are often presented with a high degree of confidence, making them difficult to identify without subject matter expertise. There is also a risk that users may place undue reliance on AI-generated outputs without sufficiently challenging the underlying assumptions, sources or conclusions. This can reduce some of the efficiency gains AI promises, as outputs often require careful review and verification before they can be relied upon.
None of these risks are reason to avoid AI. Each is a reason to adopt it with clear governance, training and oversight in place, so that the benefits are captured without exposing the charity, its beneficiaries, or its donors to unnecessary risk.
What should trustees do next?
Trustees do not need to become AI experts, but they should understand where AI is already being used within the organisation, what risks arise from that use and whether appropriate controls are in place. This includes approving an AI policy, ensuring staff receive training, considering AI-related risks within the charity's risk register, and periodically reviewing how AI is being used in fundraising, finance and beneficiary-facing activities.
The Charity Digital Skills Report is the annual barometer of digital and AI trends, funding, and support needs across the sector, including those of marginalised groups. You can view the full report here: Charity Digital Skills Report
Contact us
We're here to help - whether you have a question, need advice, or want to tell us about your requirements.