Charities must take inequality into account when developing plans to use artificial intelligence, voluntary sector leaders have warned.
In an online panel discussion at the think tank NPC’s annual conference last week, charity experts gave their views on the benefits and pitfalls of using AI to further their charitable cause.
Panellists said AI can be a useful tool for the sector, but cautioned that technology alone could not solve inequalities in data collection and analysis.
Katie Rose, director of Europe at the social policy charity the Centre for Public Impact, said: “We need to think deeply about how equity and oppression shows up in the design of AI.
“Marginalised communities don’t always get a seat at the table. AI isn’t going to change that, unless we think about existing power structures and how we dismantle them in order to ensure our data is representative – that’s how we can unlock the power of people through AI.”
Rose said it was important to have inclusive design teams when developing AI software and to make sure designers “really understand” the history of the community they are serving throughout the development process.
Laura Hamzic, director of digital communications at the sexual health and education charity Brook, added that being a “digital leader” also involves helping staff to understand AI better in order to tackle fear around the technology.
“As a digital leader, you have to help people make sense of it ahead of time, because it’s a new thing. It’s about bringing everybody on board and helping them feel a little less scared and more prepared to engage,” she said.
Brook is at the beginning stages of its AI journey, said Hamzic, and is seeking funding for its AI projects.
The charity hopes to use AI to improve access to sexual health services and tackle misinformation, by implementing an AI chatbot at its digital front door to guide users through a thorough consultation.
Hamzic hopes AI will also help identify where the charity can strengthen its service delivery.
“Using AI, we can see what questions our model was unable to answer and at what point it had to pass to a clinician,” she said. “So learning where your gaps are is an opportunity with AI.”
Neil Giles, director of intelligence and founder of the Traffik Analysis Hub at Stop the Traffik, said the charity built an extensive database using AI that captures lived experiences of human trafficking.
He said that while the system had been “highly effective”, funding these platforms can be challenging because they need “constant maintenance and upgrades”.
“You need to invest in training models that constantly develop the capabilities of AI platforms,” said Giles.