Using AI to Track Media Coverage of USAID’s Dismantling

In the June cohort of the XD131 Applied AI course, participants were asked to choose a final project that would draw from guest experts, featured tools, and practical work sessions to apply AI in a novel way to a personal interest. This post features a brief Q&A with one participant, Oliver Subasinghe, description of final project process, and then output of the final project produced for the course. 

Q&A with Oliver Subasinghe

Q: Could you share a bit about yourself and why you took a course on applied AI?
I am a former USAID diplomat who built cross-sector partnerships for climate resilience, water security, and Ukraine assistance. I also previously held a policy and data analysis role, and I was attracted to this course to update my skills as I embark on a new career path. I was also curious about how AI is rapidly evolving (for better and worse) and what the reality of these tools is versus the hype. I had also taken a TechChange-organized course in the past and found it to be highly relevant and applicable to my job. 

Q: In preparing for the final project, were there any tools, experts, or experiences that were particularly helpful?
I drew on lessons from really great guest lectures for this course, especially the speakers who shaped AI policy in their organisations. I also found the class activities helpful and not overly burdensome. The cohort group was also super fun and engaging to learn from.    

Q: Why did you choose this final project option? Did anything surprise you once you started?
I wanted to build an AI bot from the ground up, despite my limited coding skills. So, I chose the final project option using Browse.AI to scrape and organise web data from Google News based on keywords. I am also a huge news junkie, so I focused on scraping news articles related to the dismantling of my former employer. I used Perplexity to learn how to perform a media analysis, and then utilized a Google Sheets integration to pull the data into Gemini Pro for further analysis and visualization. Once I picked the source to scrape and understood what Browse.AI could do in terms of its pre-built bots and features, the rest of the experience proceeded smoothly.    

AI-generated cover image for XD131 final project

Final Project Approach

Background

The United States Agency for International Development (USAID) is an independent agency that was responsible for US foreign aid from 1961 to 2025, with average annual disbursements of about $23 billion from 2001 to 2024, missions in over 100 countries. 

Objective:

For my final project in the applied AI course, I chose an option which involved scraping and organising web data. I aimed to develop a process for scraping, monitoring, and analysing news stories about the Trump Administration's dismantling of USAID and its impacts.

Approach: 

I first consulted Perplexity to help me understand media analysis. Then, I used Browse.AI’s pre-built bot to scrape and monitor articles from Google News with the keyword ‘USAID’. Next, I used Browse.AI's integration feature to send the data to Google Sheets and promoted Gemini Pro to generate a media analysis and an infographic. The Browse.AI bot will automatically monitor more news stories and push updates to the Google Sheet every Tuesday.

The assignment took about six hours of trial and error by an amateur media analyst. I also signed up for a few trial subscriptions to access needed features (which I plan to cancel later). Nonetheless, these ‘no-code’ generative AI tools were easy to use and apply.

Visualization of AI workflow for final project

Challenges:

Several limitations and constraints arose during the process. For instance, Browse.AI limited the number of news stories I could scrape to 100 because I was using a free account. For ease of analysis, I limited my selection to articles written in English. Initially, I wanted to scrape a more sophisticated news aggregator called the USAID Media Coverage dashboard. My free version of Browse.AI was unable to scrape that site. So, I turned to the pre-built bot option for extracting stories from Google News using keywords instead.  Google Pro also couldn't perform a ‘deep dive’ review of the article content and only provided headlines and URLs for analysis. 

Conclusions:

Overall, I was amazed at how much I could accomplish with what is a complex task in just a few hours, using some free accounts and the excellent sessions and instruction from our Introduction to AI course. With a budget and more time, I thought of doing a deeper dive on the article content and localising the news sources in countries most exposed to the cuts in U.S. foreign assistance. This experience gave me a great deal of respect for the social scientists and professionals who analyse media sources, which is a much more complex and nuanced endeavour.

Results: 

Below is a summary of the media analysis and visuals. You can also view all the project outputs using the links below:

• Full Media Analysis and Infographic (Perplexity; Gemini Pro)

• Extracted News Articles Data (Google News; Browse.AI; Google Sheets)

Final Project: USAID in the News: Funding, Politics, and Global Reach

News coverage of USAID predominantly focuses on funding and program changes, with frequent mentions of "cuts," "aid," and "funding," often alongside discussions of programs potentially ending or being dismantled.1 Political figures like Trump and Musk are central to these narratives, indicating their significant influence on USAID's operations.2

The sentiment of the coverage is largely neutral to positive, with negative reporting being minimal, typically surfacing around budget cuts or controversies. Top sources like AP News, The New York Times, and NPR provide broad and diverse coverage, reflecting a mix of reliable and some politically leaning outlets.

Spikes in coverage occurred in June and February 2025, suggesting key events drove media attention during these months. Geographically, Sudan, Russia, Colombia, Mali, and India are the most mentioned countries, implying that news often discusses USAID's programmatic changes and political influences within these nations, highlighting potential impacts on their aid and development initiatives.