AI Tools Help Grassroots Groups Monitor Human Rights Violations

Rahul Bhargava, a professor of journalism and art/design at Northeastern University, has developed AI tools to assist grassroots organizations in monitoring human and civil rights violations. These tools are designed to compile incident data from media reports, addressing issues that have a significant impact on society.

Real-World Applications

Examples of incidents monitored include a shooting in Atlanta, a racially motivated assault in Bellingham, Washington, and a case in Ohio where a woman faced felony charges following a miscarriage. These incidents, reported by local media, are part of a broader pattern of civil and human rights violations that are often not officially monitored.

Grassroots Efforts in Data Collection

Grassroots organizations play a crucial role in gathering information from media reports about alleged violations. These include gender- or identity-based killings, police killings, white supremacy-motivated attacks, mass shootings, and pregnancy criminalization.

Collaboration with Catherine D’Ignazio

Bhargava and his collaborator, Catherine D’Ignazio of MIT, are using AI to monitor these violations. Their AI tools have significantly reduced data extraction times for partners, from six hours to one hour, making the process more efficient and less burdensome.

Previous and Ongoing Projects

Bhargava and his students have previously developed AI tools for the Data Against Feminicide project, which monitors news on alleged intentional killings of women in North and South America. They have collaborated with organizations like the Sovereign Bodies Institute, African American Policy Forum, and Feminicidio Uruguay.

New Grant and Expanded Efforts

With a new grant from the Data Empowerment Fund, Bhargava is managing interactions with a broader set of grassroots groups across the U.S., Mexico, Uruguay, Brazil, and Argentina. Each collaboration involves designing custom workflows and creating specific articles to train the AI model.

Custom AI Models for Specific Needs

“We build a new model for every group we work with,” Bhargava explains. Each group’s nuanced definitions of the types of cases they care about necessitate custom AI tools. The models combine media databases, machine learning algorithms, and email alerts to report possible cases.

How the AI Tools Work

The AI model ingests large sets of news stories from archives like Media Cloud and NewsCatcher, filters them by keywords, and categorizes the data using a custom AI classifier. The model checks archives three times a day, delivering results via email and allowing real-time monitoring through a live dashboard. It clusters similar stories, reducing the need for activists to read all reports.

Benefits of the AI Tools

These tools enable grassroots groups to efficiently gather data, raise awareness, provide legal help, and advocate for policies to protect vulnerable populations. Bhargava notes that in many countries, such data isn’t officially collected, making these efforts crucial.

Open Source Code

The code developed by Bhargava is open source, meaning it is publicly accessible and available for modification or distribution to other interested groups. This openness fosters collaboration and wider use of the tools.

Addressing Data Collection Challenges

Human rights grassroots groups often spend hours reading news and compiling stories, a tedious and emotionally taxing process. Bhargava’s AI tools streamline this process, offering a more effective alternative to inadequate tools like Google News Alerts, which often return irrelevant stories.

Conclusion

Bhargava's AI tools represent a significant advancement in the ability of grassroots organizations to monitor and address human and civil rights violations. By leveraging AI, these groups can more effectively compile and use data to advocate for vulnerable populations and drive social change.