So excited to announce that we’ve received a grant to support our work on the Counterdata Network from the Data Empowerment Fund. Our project builds on a powerful legacy—think of Ida B. Wells using news to expose lynchings in the late 1800s. Today, we’re continuing this tradition by developing AI models that empower advocates to uncover human rights violations more effectively. This involves custom-trained machine learning models to identify relevant events, designed with a focus on equity, localism, and data justice, all using a deeply human-centered approach. Our software tools are actively used by dozens of human rights groups. With this support, we’ll be able to improve their stability and expand our network of partners.
Pushing Tech Forward
Our system pulls from various online news archives, both open-access and subscription-based, using keywords and dates to fine-tune search. We’re also developing models leveraging techniques like transfer learning and word embeddings and offer browser extensions that integrate directly into advocates’ workflows.
Rethinking the Model
Traditional commercial AI doesn’t meet the needs of these groups, so we’re adopting a co-creation approach, rooted in feminist principles. We’ve spent four years building trust and interviewing community members—often taking months before the first line of code is written. This focus reduces emotional strain and widens the scope of what can be achieved.
Open Questions
Many questions remain for us: How do we sustain these civic technologies as public goods? What community governance models make sense? And how do we best handle data from social media monopolies? Our hope is to contribute not just tools but also new ways of thinking about AI that align with human rights and justice.
Read more at Northeastern Global News or on the Data Against Feminicide website.