With the rapid proliferation of daily developments in tech advances, we explore whether charities are applying AI and data science in an ethical way.
There are cool, new AI and data science developments appearing every day. Such advances lead to the all-important question: are charities applying AI and data science in an ethical way?
As digital tech evolves in leaps and bounds, the need for regulation emerges. Charity Digital Code chair Zoe Amar states that charities need to be aware of any potential harm that can come from tech used inappropriately and to safeguard service users from risks. The Charity Digital Skills Report suggests that the ability to assess how responsible your charity is with digital is now considered a core skill. There are concerns around transparency, fairness and consent.
DataKind UK worked to pinpoint significant ethical principles to gauge the ethical impact of data projects. They worked with the Association of Medical Research Charities (AMRC) to devise a framework to appraise the many codes in existence and find the ethical principles most applicable to them. They reviewed the wide range of guidance available and discussed, in depth, the issues and concerns of AMRC and their members.
They identified nine key principles:
Beneficence: focus on benefiting people
Nonmaleficence: avoid harming people
Autonomy: empower people to make informed choices
Justice: ensure fairness in distributing benefits and risks
Explicability: give reasons for the outcomes of algorithms
Open research: make research accessible and open for reuse
Community-Mindedness: commit to collaborate with others
Proportionality: balance the potential risks and benefits
Sustainability (financial and operational): Reduce the harm of developing digital tech and identify sustainable options
Although DataKind UK focused on health charities, their approach works for all types. Having identified your ethical principles, you need to think about how to apply them. DataKind UK produced questions to discuss with tech partners as a result of discussion with AMRC members.
The BBC hosted a discussion about the challenges charities can face when using AI for social good.
Giselle Cory, Executive Director at DataKind UK described how DataKind, in addition to other projects, had helped make a predictive model to pinpoint people who might need extra support whilst waiting in line for the food bank. Although she advised that relying only on AI-based decisions can be tricky in light of ethical concerns.
Julie Dodd, Director of Digital Transformation and Communication at Parkinsons UK, explained that Parkinson’s UK are working with Benevolent AI to help discover medical solutions by scanning millions of Parkinson’s-related papers to spot medical clues missed by humans.
Michelle Eaton, Programme Manager in the Government Innovation Team - Data at Nesta, revealed that Nesta are using AI to research wider problems and learn more to find possible solutions. She said that “an AI solution is not a silver bullet” and it is important to gather data before making decisions.
The panel addressed possible ethical issues around using AI and mentioned the importance of strong ethical frameworks to ensure good practice.
Our podcast episode, titled Safeguarding the sector: designing digital ethics for your charity, is about the responsibility of charities to uphold and demonstrate ethical behaviour to gain the general public’s trust. This episode focuses on digital, data and technology and considers how any innovation can have unpredictable outcomes.
Charities must evaluate their use of tech, such as AI, and review how it affects their mission to avoid harm and support progress, particularly when challenging existing biases in society.
We reported that experts believe AI will shape society in the next few decades but has its own problems such as concerns over data use, loss of jobs and more. Charities are using machine learning, chatbots and more to provide information, process voice-based donations and drive innovative research.
Our podcast episode, titled Is ignoring AI the charity sector’s biggest tech mistake?, discusses this issue with experts in the charity sector. It covers topics such as processing data through algorithms, how AI has raced ahead and uses the phrase ‘Responsible AI’ as a key term to emphasise doing beneficial things and helping people. They refer to the importance of using human judgement and discernment with real-life examples. The experts argue that coding has become easier, so AI can become easier.
From large, well-known charities to smaller ones operating locally, applying AI and data science in an ethical way is crucial for everyone in the charity sector. Luckily, you are not alone and can lean on the tech expertise, guidance and experience of charities already embracing such advances. Slick, smooth tech can seem daunting but with a bit of practice, and info, you can use these tools to better meet your mission. You can enhance from your current use of digital tech to go to infinity and beyond!