Strategy, skills and shared standards key to making AI work for philanthropy

Most Italian foundations are using artificial intelligence (AI), but in almost all cases it is done in an informal rather than strategic way, a new report says.

The researcher behind the report, Giulia Togna from Milan’s Università Bocconi, says she hopes it will support philanthropic organisations to adopt AI “in a way that is informed, responsible, and consistent with its potential transformative role”.

Togna’s research was commissioned by Italian philanthropy sector body Assifero, and is based on responses from 33 members of the organisation in late 2025 and early 2026.

Experimentation without strategy

It shows that 88% of the 33 are using AI, but only eight have an internal policy on AI use, and only one organisation has reached what Togna termed “strategic adoption”.

The other organisations in the study were evenly split between “explorative adoption” and “structured adoption”, with larger organisations more likely to be in the latter, more advanced phase than smaller ones.

The research said that the organisations most likely to be in the structured adoption stage were corporate foundations. Togna says that sharing the infrastructure and skills of their corporate patron was one likely reason for this.

Between half and two thirds of respondents are already using AI for research and analysis, and production of written content or communications.

In addition, many respondents saw the prospect of it being used for reporting, strategy, monitoring and impact measurement in future — although only one or two respondents are currently doing this.

Skills gaps the biggest barrier

Asked about the barriers to further adoption of AI, the most common responses were:

• Internal skills gap (58% of respondents)
• The lack of a strategic vision for the implementation of AI (45%)
• Concerns of diluting the organisation’s human touch (38%)
• Resource concerns (35%)

In her conclusion, Togna says that Assifero could support the uptake of AI within philanthropic organisations by “fostering spaces for peer learning, critical dialogue, and the exchange of diverse experiences, thereby helping to drive change processes aligned with the values ​​and unique characteristics of the philanthropic sector”. She ends:

“This study represents a first step in that direction: a starting point for ongoing reflection — conducted alongside foundations and philanthropic entities — on how to guide the adoption of AI in a manner that is informed, responsible, and consistent with its potential to drive transformation.”

Fundraising Europe readers will recognise the pattern of AI already being implemented, but in an unplanned manner. The same picture was clear in two recent reports — one published by France générosités ​in June 2026, and another by tech supplier Blackbaud, focusing on UK charities.