The Chartered Governance Institute UK and Ireland is warning that charities are embedding artificial intelligence technology faster than their governance systems can effectively oversee its use.

Its survey of trustees found that more than nine in ten charity board do not even have a “formal AI kill switch policy” in place.

This is a technical safety mechanism designed to shut down or isolate an AI system when it behaves unpredictably or malfunctions.

But only 5 charities involved in the research have such a policy in place, compared to 68 who do not.

Also, the Institute found that almost two in five of charities are not using a human to audit AI generated content.

Just under a half have no policy on managing the retention and deletion of raw AI-generated meeting data and almost three in four have not conducted a third-party AI risk assessment.

In addition, only around half of charities are able to provide evidence to substantiate AI claims in public disclosures.

“Our research found broad agreement across sectors that AI adoption is advancing faster than organisations’ ability to govern it effectively,” said report author and policy advisor to the Institute Dr Valentina Dotto.

“Many boards recognise the opportunities AI presents, yet they often have limited visibility over how these systems are being used across their organisations. Governance frameworks, oversight structures and board capability are not developing at the same pace as adoption.

“Boards need clear oversight of where AI is being used, how decisions are being influenced and what controls are in place to ensure outcomes remain explainable and accountable.”

The Institute is urging organisations to strengthen visibility over AI use, embed AI within governance and risk frameworks, improve board oversight capability and ensure accountability remains clear wherever AI is deployed.

Governance concerns

This is the latest evidence to emerge raising concerns around the effectiveness of charity governance to meet the escalating use of AI in the sector.

This year’s Charity Digital Skills Report found that just under half of charities are not taking action to manage risks and progress with AI.

Also, while almost four in five large charities have, or are developing, an AI policy, this proportion falls to only around one in three among small charities, this report found.

It also revealed that governance around AI is slipping down boards’ agendas.

Less than one in five are reviewing data protection, security and safeguarding around the technology, down from more than one in three last year.

Strengthening digital data, AI governance and decision making is only a priority among a third of charities, although this proportion doubles among large charities.

Damage to public trust in civil society, unless charities improve their governance around AI, is a major concern raised by the Charity Excellence Framework in a report it released in April.

It warned that while the public is not against charities using AI, they are still “cautious” about its use, and that trust in the sector’s responsible deployment of the technology is “conditional”.

---