Artificial intelligence (AI) is proving to be a lasting presence in our lives. The rise of generative AI tools, such as Claude, Copilot, and ChatGPT, has made these technologies a staple on our desktops. AI is increasingly woven into the fabric of our day-to-day activities, influencing decisions like the advertisements we see on social media or our visa applications. The implications are particularly significant for the communities that charities exist to serve.
Charities Can Shape AI Without Technical Expertise
Charities focused on health, social care, or environmental issues have historically monitored AI developments without direct involvement in technology. However, as AI becomes more prevalent in these areas, there’s a risk that charities could fall behind, further marginalizing vulnerable communities in an AI-driven world.
Guide AI Data Practices Relevant to Your Cause
AI tools rely on data to be functional. Neglecting to responsibly consider data in the context of AI is akin to admiring a car in a showroom while ignoring its need for fuel.
Charities are custodians of invaluable information about underrepresented groups and have the opportunity to advocate for their inclusion in the development of ethical and inclusive AI models. As the World Economic Forum notes, this stage in the AI life cycle offers a crucial chance to incorporate inclusivity, as the data forms the bedrock of the model.
Civil society also plays a significant role in data governance, ensuring high-quality, representative data is collected and safeguarded with appropriate protocols.
Ensure AI Design is Inclusive
User acceptance testing is a vital step in the AI development process where charities can have a meaningful impact. The voluntary and community sectors possess extensive experience in inclusive service design and co-production, which can significantly influence AI development. Examples like Microsoft’s inclusive design principles can be enhanced by the social sector’s expertise.
Charities engaging in testing or shaping the testing process can foster better governance for AI models, ensuring these technologies are used ethically and effectively serve all users, especially those from marginalized backgrounds.
Get Involved in AI Governance
AI regulation in the UK is still evolving. This presents a key opportunity for charities to amplify the voices of the communities they represent as discussions turn into legislation and governance frameworks.
Research from the Ada Lovelace Institute highlights a significant gap between public expectations and governmental aims regarding AI. One recommendation from the report is to establish formal channels for civil society, especially those representing vulnerable groups, to contribute meaningfully to regulatory processes.
How Charities are Shaping the Future of AI
Certain organizations within the charity sector are proactively engaging in technology dialogues with their peers. For instance, DataKind facilitates discussions and provides a variety of resources on the topics of AI adoption, policies, and ethical considerations.
Thanks in part to advocacy from the Public Law Project about AI use in government departments, the government’s Responsible Technology Adoption Unit has introduced an algorithmic transparency standard that mandates the publication of models significantly involved in public decision-making.
CAST has launched the Charity AI Task Force to accelerate the social sector’s adoption of AI in line with the commercial sector, while also serving as a liaison for the government to tap into the vast skills and community insights the sector offers. In 2025, the government pledged to reset its relationship with civil society via the launch of the civil society covenant—an additional means for social sector expertise to influence AI policy and practice.
Charities are also collaborating directly with AI developers to create positive impacts. For example, The Children’s Society partnered with Microsoft to create an AI tool tailored to support young migrants and refugees.
While strides are being made to include charities and marginalized groups in the evolving landscape of AI, much remains to be accomplished. The risks of exclusion are significant and may compound over time. Nevertheless, there are promising opportunities for both the charity sector and AI developers to foster an inclusive, representative, and ethical approach to AI, ensuring that these technologies support charitable missions effectively.
