Risks and rewards of AI for charities

  • Artificial intelligence (AI) is no longer on the horizon – it is already being applied widely across the public and voluntary sectors
  • For charities, AI offers wide-ranging benefits, including new ways to deliver services and engage with donors and fundraisers
  • We look at the opportunities AI presents, and the risks charities must negotiate

It feels as if every week there are new stories about artificial intelligence (AI) being put to innovative use in the public sector – from marking school essays to diagnosing illnesses.

But how could it affect how charities operate? Here, we consider some of the most significant benefits of AI for voluntary organisations, and some of the associated challenges.

Donor and fundraiser engagement

AI and other Fourth Industrial Revolution technologies, such as Robotic Process Automation, are helping charities engage more effectively with their donor bases.

Examples of this include AI-driven chatbots and virtual assistants; many of the UK’s larger and more established charities have already taken steps in this area. The Children’s Society has trialled a chatbot application on its Facebook page to answer fundraising questions, while Arthritis Research UK has a virtual personal assistant.

AI can also play a direct role in driving donations. The NSPCC, for example, now enables donations to be made via Alexa.

Transforming delivery models 

AI enables charities to make the most of limited resources in order to deliver their social mission. AI can analyse large data sets and spot patterns far more effectively than humans can. This can help charities when making decisions about when and where to allocate resources to services users.

For example, charities that support people with physical or mental health conditions, or individuals receiving drug or alcohol rehabilitation, could use AI to determine the most effective programme for each service user.

But it is not just behind the scenes that AI could potentially reshape the way services are delivered. AI-powered virtual assistants could also play an important role – for example as life-aids for individuals with physical disabilities, or voice-based assistants providing guidance and advice to those dealing with homelessness or older people living alone.

Improving insight and analysis 

AI’s ability to spot patterns in large and complex datasets can generate valuable insights for charities. This could include:

Fraud, in particular, is a key challenge for both the charitable sector and the insurance industry. Zurich uses a technological solution called NetReveal, which combines elements of predictive analytics, social network analysis and machine learning, to uncover hard-to-detect suspicious behaviour.

Should charities have concerns about AI?

While there are wide-ranging benefits, charities should be aware of the potential risks when embracing AI. One of the key challenges, according to Gordon Wilmott, Head of Charities and Social Organisations, Zurich Municipal, is trust.

Charities must be clear about what kind of data they are collecting in order to power their AI, and how it will be used. The public must also have confidence their data is being looked after properly.

Gordon says: “It is vital that charities maintain the confidence of service users. How they look after and use their data will directly correlate with how willing people are to accept the advances charities want to make with AI.

“Indeed, under GDPR, organisations must be able to explain how they reach certain algorithmic-based decisions about their customers or service users. In order to meet this duty, it is important that charities are able to explain the workings of any AI they deploy, and that there is transparency.”

Research reveals concerns of AI impact on vulnerable groups

Charities considering using AI to support vulnerable groups should also be prepared to meet with some opposition.

Our latest whitepaper, Artificial intelligence in the public sector: the future is here, reveals that the public has considerable reservations about AI being used to provide services in this way. Only 16% of people approve of AI being used for the care of the elderly, and only 9% of approve of its use with vulnerable children. These are two areas which are already in significant development, particularly by local authorities.

Charities will need to think carefully about how their activities will be perceived by staff, supporters, donors and service users.

“If charities damage that bond they will find it incredibly difficult to engage with these groups and to deliver on their social mission using AI,” says Gordon.

Taking a holistic view of AI

For Gordon, it is important that charities look at the bigger picture, and not just the individual benefits of each new piece of technology.

“Charities need to look at their systems as a whole, and ensure they are putting in place the appropriate governance frameworks for any AI that they deploy. How do they create the right QA frameworks and recruit and retain trustees with the right skillsets and experience?” he adds.

Our new report, Artificial intelligence in the public sector, takes an in-depth look at the ethical and governance challenges associated with AI. It also provides clear, practical guidance to help organisations, including charities, understand the questions they should be asking when embarking on AI projects.