The third sector needs better infrastructure, not more prompts
The AI hype is reaching fever pitch, and the third sector is no exception. According to the 2026 Charity Digital Skills Report, 79% of charities are using AI – a jump up from 76% in 2025, 61% in 2024 and 35% in 2023.
When we look a bit deeper, the story gets more complicated. Whilst AI usage is high, only 28% of the charities surveyed report having a digital strategy in place. In 2024, this number was 50%. This begs the question: Do charities have the kind of technical infrastructure and governance in place that’s needed for AI use to be effective?
The latest tools won’t help if infrastructure is missing
Keeping up with AI is a tall order. When you’re spread thin working on mission-critical stuff with minimal resources and budgets, it gets even harder. And the truth is, AI is at its most effective when it’s integrated with robust technical infrastructure.
The power and the rate of change in AI capabilities are truly breathtaking, but building a skyscraper on shaky foundations is high risk. AI can automate and integrate systems and processes, but if the underlying data quality is poor or systems are not well configured, this may create more problems than it solves. Turbocharging poor or inaccurate processes will simply make things worse, faster.
– Alistair Forbes, Founder and CEO, Scottish Tech Army
As Alistair suggests, you can’t build a skyscraper on a foundation of messy spreadsheets and legacy software. Existing research highlights the following as critical factors for AI success:
Data reliability: A 2026 report from Informatica cites unreliable data as the top barrier to success for AI projects – and yet 65% trust AI outputs even when data gaps remain. If data isn’t clean, integrated, and fit for purpose, then it isn’t AI-ready.
AI governance: AI usage is increasing faster than the development of rules, practices, and policies that ensure it’s being used responsibly, ethically, and legally. This problem exists everywhere but is a point of emphasis for charities especially.
AI confidence and literacy: Half of charities say that keeping up with emerging tech trends is a challenge – 41% say they’re poor at it and 9% don’t do it at all. On top of that, 35% say they are poor at using AI tools in their everyday work and 29% don’t use them at all.
If these kinds of precursors are missing, it’s almost impossible for an organisation to solve meaningful problems with AI.
Citizens Advice's CADDY is a great example of developing an AI tool that has been built on a foundation of trusted data sources, robust governance and user experience. It's transformative, enabling the organisation to support more people with the same resource. –
- Kirsty McIntosh, Executive Director, Scottish Tech Army
You can’t automate chaos
We’ve seen first-hand how this gap between hype and reality can play out. Charities may feel under pressure to adopt the latest tech, but the real impact usually begins when you fix invisible friction and start creating technical infrastructure that will scale. Without this, there’s a danger of automating chaos rather than solving problems.
Ultimately, this is the kind of thing the Scottish Tech Army and the Tech for Good Alliance are here to help with. If you’re a third sector organisation facing problems that require technical or digital solutions, AI-related or otherwise, contact us through our request form.
