AI Resilience Report forPolitical Scientists

Right now, AI is mostly augmenting political scientists rather than replacing them — but the tools are getting more powerful every month. An APSA Presidential Task Force chapter, using a task-based framework from labor economics, analyzed which political functions AI can substitute for, leave unchanged, or augment, finding that in election administration adoption remains limited because of AI's tendency toward hallucination, restricting it to back-office tasks. On Capitol Hill, the Congressional Research Service faces a significant backlog of bill summaries as more than 15,000 bills have been introduced in the current Congress, and a House Administration Committee hearing is examining whether AI-enabled policy analysis can help CRS modernize its research infrastructure — with Acting Librarian of Congress Robert Newlen requesting $5.4 million to build a generative AI platform to summarize legislation faster ^[1].

For polling, Pew Research Center's methods lead, Courtney Kennedy, explains ^[2] that some firms now practice "silicon sampling," asking AI what the public would think, though studies have found that AI estimates tend to stereotype groups of people, have a harder time representing Republican viewpoints than Democratic ones, and understate the level of disagreement in public opinion. Teaching is also shifting: a Journal of Political Science Education study asks whether instructors can design assignments that reduce inappropriate AI use while encouraging positive learning uses ^[3].

Adoption is uneven. Cheap, commercial chatbots make it easy to summarize policy documents, code survey data, and draft memos, and Research.com's 2026 outlook on political science careers notes AI is rapidly reshaping tasks like data analysis and legislative research ^[4]. But trust matters here: Pew flatly refuses silicon sampling on ethical grounds, and hallucination risks slow rollout in elections and government.

Brookings researchers also point out that AI exposure has become a politically charged issue itself ^[5], which makes public-sector adoption slower and more scrutinized than in private industry. The good news for you: judgment, ethics, teaching, and advising students — the highest-value parts of this career — are exactly what AI struggles with most.