79% of charities now use artificial intelligence, few use it strategically. What the Charity Digital Skills Report 2026 means for your CRM and your data.
The Charity Digital Skills Report is the closest thing the United Kingdom voluntary sector has to an annual health check of its digital life. The 2026 edition, published in July by Zoe Amar and Nissa Ramsay with responses from 807 charities, their largest response yet, contains one headline number and a set of quieter ones that matter more.
The headline: 79% of charities now use artificial intelligence routinely. Adoption is no longer the story. What organisations do with it is, and that question runs straight through your customer relationship management (CRM) system, which is why we wanted to write about the report here.
The quieter numbers
79%
of charities now use artificial intelligence routinely, yet only 28% have a digital strategy
Read past the headline and a pattern emerges. Progress is real: 81% of charities say they made digital progress in the last year, up from 63% the year before. But look at what the artificial intelligence is being used for. Around 60% use it for routine tasks such as emails and meeting notes, 47% for research, 43% for idea generation. Just 9% are using it to change how services are delivered.
Meanwhile, only 28% of charities have a digital strategy at all. Just over half rate themselves fair or poor at using data to make strategic decisions. And 56% name limited skills as their biggest barrier, with a third of trustees and a quarter of chief executives rated poor on artificial intelligence skills. Among small charities, only 36% have an artificial intelligence policy in place.
In short: nearly everyone is experimenting, very few are being strategic, and the sector knows it. As one funder quoted in the report’s launch coverage put it, if any other tool had 79% uptake with one in three people unable to use it effectively, we would have done something about it by now.
Why this is a CRM question
The distance between experimenting and strategy is not a licence purchase. It is data.
Drafting an email with artificial intelligence needs no infrastructure at all, which is why 60% of charities are already doing it. Using artificial intelligence to spot supporters at risk of lapsing, to prepare a caseworker’s morning briefing, or to answer a donor’s question accurately needs something else entirely: supporter and participant data that is complete, current, deduplicated and held somewhere the technology can safely reach it. That place, for most charities, is the customer relationship management system.
This is the uncomfortable dependency the report’s numbers point at. The 51% of charities that rate themselves fair or poor at using data for decisions cannot fix that with a chatbot, because the same scattered spreadsheets and duplicate records that undermine human decisions undermine artificial intelligence ones, only faster and with more confidence. Strategic use of this technology is downstream of a data foundation. There is no route around that; there is only the work.
The encouraging sign in the same report: 51% of charities now name data collection and analysis as a priority, up from 37% last year. The sector has spotted the dependency too.
Four moves that turn experimenting into strategy
This is the path we walk with charities, and none of it requires buying anything this quarter.
Fix the foundation. Pick your single source of truth for supporter and participant data and consolidate towards it. Deduplicate, standardise, and retire the side spreadsheets. Every future use of artificial intelligence gets better or worse depending on this step, which is why it comes first.
Write the policy before you need it. A one-page artificial intelligence policy, covering what tools are approved, what data may never be pasted into them, and who decides, moves you out of the risky 64% of small charities without one. It also gives trustees, a third of whom the report rates poor on these skills, something concrete to govern.
Choose jobs, not tools. “We should be using artificial intelligence” is not a plan. “We want to halve the time fundraisers spend preparing donor briefings” is. Name two or three jobs with measurable outcomes, check your data can support them, and pilot those. Everything else can wait.
Budget for people, not just licences. 59% of charities now name upskilling staff as a priority, up from 43% last year, and 44% name training as their greatest funding need. A modest tool in trained hands beats an expensive one in confused ones, every time.
Where this goes next
The report’s authors have called on funders to move fast on skills and infrastructure funding, and the full report is worth your leadership team’s time; Zoe Amar Digital carries the details. Our contribution is narrower: making sure the data foundation those ambitions rest on actually exists, in Salesforce, for the charities we work with. It is the same foundation that pays off in unglamorous ways long before any agent arrives, from reclaiming Gift Aid properly to sending supporters email they actually want.
If your organisation is somewhere in the 79%, experimenting and wondering what strategy would look like, book a discovery call. Bring your messiest data problem; we have seen worse.