AI Agent Operational Lift for New America in Washington, District Of Columbia

Why AI matters at this scale

New America, a Washington D.C.-based think tank founded in 1999, operates at the intersection of research, policy advocacy, and public engagement. With 201–500 employees, it sits in a mid-market sweet spot: large enough to have structured workflows and a national reputation, yet agile enough to adopt new technologies without the inertia of a massive enterprise. The organization produces a high volume of written content—policy papers, op-eds, grant proposals, event summaries—and manages relationships with donors, fellows, and policymakers. These are precisely the kinds of text-heavy, knowledge-driven processes where generative AI can deliver immediate, measurable gains.

At this size, AI adoption is not about replacing human expertise but amplifying it. Analysts spend significant time on literature reviews, data gathering, and drafting. AI can compress these tasks, freeing intellectual capital for deeper analysis and creative problem-solving. Moreover, the think tank’s reliance on grants and donations makes efficiency and demonstrable impact critical. AI-driven donor intelligence and automated reporting can directly strengthen funding pipelines.

Three concrete AI opportunities with ROI framing

1. Research acceleration and synthesis
Policy researchers often sift through hundreds of pages of legislation, academic papers, and news. A retrieval-augmented generation (RAG) system over internal and external databases can summarize, compare, and extract key arguments in minutes. ROI: conservatively saving 10 hours per researcher per week, allowing the same team to produce 20–30% more output or deepen analysis quality.

2. Grant writing and reporting automation
Drafting proposals and funder reports is repetitive yet high-stakes. Fine-tuning a large language model on past successful submissions can generate first drafts that human writers then refine. This reduces turnaround time from days to hours. ROI: higher win rates and more time for relationship-building, potentially increasing grant revenue by 5–10%.

3. Donor engagement personalization
Using machine learning to segment donors and predict interests enables tailored communications and event invitations. Automated sentiment analysis of donor interactions can flag at-risk relationships. ROI: improved donor retention and average gift size, with minimal additional staff cost.

Deployment risks specific to this size band

Mid-sized think tanks face unique risks. First, data sensitivity: policy research often involves confidential sources or pre-publication data. Any AI system must be deployed with strict access controls and preferably on a private cloud or on-premises to avoid leaks. Second, reputational risk: an AI-generated error in a public-facing report could damage credibility. A human-in-the-loop validation step is non-negotiable. Third, change management: with 200+ staff, there will be varying levels of tech literacy. A phased rollout with champions in each program area is essential to avoid resistance. Finally, cost predictability: while AI tools are increasingly affordable, uncontrolled usage can lead to budget surprises. Governance around API consumption and license allocation is needed from day one.

By starting with low-risk, high-ROI pilots and building internal AI literacy, New America can turn its size into an advantage—moving faster than larger institutions while having the resources to implement robust, secure solutions.