We are living in an extractive data economy. Every day, people generate a firehose of new data on hundreds of apps and services. These data are often sold by data brokers indiscriminately, embedded into user profiles for ad targeting, and used to train large language models such as ChatGPT. Communities and individuals should benefit from data made by and about them, but they don’t.
Related Content
That needs to change. A report released last month by the Aspen Institute, where I work, calls on foundations and other donors to lead the way in addressing these disparities and promoting responsible uses of data in their own practices and in the work of grantees. Among other things, it suggests that funders encourage grantees to make sure their data accurately represents the communities they serve and support their efforts to make that data available and accessible to constituents.
I’ve seen first-hand how philanthropy can effect huge change in this area. As CEO of Creative Commons, a nonprofit that promotes greater access to content, I was part of a broad coalition of organizations that advocated for expanding the availability of research, data, and academic papers. The academic publishing industry often restricted access to and resold content produced by researchers and funded by philanthropy and governments.
We asked funders to modify grant agreements to require that the research they supported was published under terms of open access. The Gates Foundation was an early adopter, asserting that if grantees take a dollar of their money, published research had to be available to all. Other foundations, including the Hewlett, MacArthur, John S. and James L. Knight, and Open Society foundations and the Wellcome Trust, soon joined. Even the federal government adopted policies related to open code, open research, and open data, putting more research in the hands of the communities that need it, not just those that can afford to pay.
(The Hewlett, MacArthur, and Open Society foundations are financial supporters of the Chronicle of Philanthropy.)
But much more needs to be done. The sheer volume and types of data generated by smartphones and other connected devices — and mined and monetized, often without people’s consent — demands an even greater philanthropic response.
Such data are now used to train A.I. models that drive decisions in areas such as access to banks and banking services and hiring. The result is a concerning lack of balance: Too few people control and capitalize on the data of the many, while individuals, communities, and society as a whole are left out. A lot of policy discussion and media coverage rightly focus on how that data is used to harm and discriminate, but more attention could also be paid to how it might help and empower communities — if they can access it.
Promoting Data Equity
Ensuring that communities benefit from data was the goal of the Council for a Fair Data Future, a group of academics, technologists, policymakers, and advocates assembled by the Aspen Institute to develop the new report. Community in the context of data can mean many things, including geographic area, ethnic or cultural group, individuals and families with specific medical conditions, or users of a common platform or service.
Given philanthropy’s outsize influence on programs that help communities, the council identified foundations as particularly well-suited to expand data access and equity — and curtail the concentrated power of corporations and governments over such data. Here are four key strategies grant makers can deploy.
Establish foundational questions about data. All grant applicants and grantees should be able to answer how and why they collect and store data on any program and its participants, how the data will be used, and what their plans are for sharing it publicly.
They should also be able to explain how they’ll ensure suppliers and partners with access to that data, use it in ways that benefit communities rather than harm them, while maintaining their privacy and security. A nonprofit might develop a plan, for example, that allows a local government to use its data to support the expansion of a youth parks program, but not as a tool for racial profiling.
Many nonprofits are already doing this planning in an ad-hoc way, but in some cases answering these questions might prove onerous to grant applicants and grantees. Funders can help by providing applicants with clear guidelines about what they expect and what’s needed. Once a grant is awarded, they can offer additional funding, training and technology to devise ongoing data equity plans, and help grantees shoulder the burden of these requests.
Encourage better data stewardship. Grantees should assess how their data is stored, distributed, secured, and managed. This will protect the privacy of groups the nonprofits serve and ensure only organizations that want to help them have access to their data.
Grant makers could also provide funds and training to grantees and their partners who want to collect and use data to create collaborative organizations such as data trusts or cooperatives. Under data cooperative arrangements, nonprofits work together to protect their data, while a data trust is an outside group that manages data on behalf of nonprofits.
Both types of organizations evaluate how nonprofit data is used and make sure it doesn’t end up in the hands of those who might deploy it for harmful purposes. These collaborative groups also help people such as academics and community advocates more easily access the data for their own work.
Make data available for community use. This would allow specific groups to refine and categorize data in ways that better represent their reality, which would lead to more representative datasets, A.I. models, and products. For example, the Asian American community is made up of more than 20 racial and ethnic groups, but often the data collected from each is lumped together with no way to distinguish among them.
At the same time many communities still only orally pass down vital information about their cultures to the next generation, which means some data is lost. Grant makers should fund efforts to help communities track down and archive this type of information.
For example, our report cites the work of social demographer Desi Small-Rodriguez who travels to different tribal communities and works with Indigenous people to identify and collect data that is important to them, including “language repositories, health assessments, demographic and economic surveys, and even fish counts.” Efforts like this, which give tribal communities more control over and access to their data, would benefit from philanthropic support.
Teach communities how to collect, manage, and use data. Grant makers should fund efforts to provide data training to local leaders. Many are doing their best to manage data collections about the programs and communities they serve but often don’t have the time or staff to refine their techniques. Trusted local institutions such as libraries are increasingly becoming venues for skills training in areas such as computer literacy and programming. These institutions could also provide data management training, including coding, cybersecurity, and more, potentially opening up opportunities for career advancement and financial stability for participants.
Responsible data stewardship comes with ongoing maintenance costs. But funders don’t need to change what they do to support this work. They can embed data equity, education, and community engagement into everything they fund — as well as their own practices. It’s the best way to ensure a fair data future.