Minnesota AI App Aims to Boost Civic Engagement Through Legislative Translation
Minnesota AI App Boosts Civic Engagement via Legislative Translation

Minnesota AI App Aims to Boost Civic Engagement Through Legislative Translation

For many residents of Minnesota, the intricate processes of state legislation can seem deliberately confusing, creating a significant barrier to civic participation. A new application called CivicLoon, developed by a local software engineer, seeks to bridge this gap by using artificial intelligence to translate and summarize legislative activities for the public.

Developer's Motivation and Background

Colin Lee, a software engineer from the Lakeville area and a former DFL candidate who ran multiple times against Republican incumbent Mary Liz Holberg for a state House seat, created CivicLoon. Lee was inspired by his political campaigns, where he observed widespread voter confusion about candidates and policies.

"I would ask people about the race and they would be misgendering my opponent when they're telling me who they're going to vote for... you just knew they had no clue," Lee recounted. Despite never winning the seat, he continued running on principle, believing that challengers in long-shot races can force important conversations on substantive issues like healthcare.

Wide Pickt banner — collaborative shopping lists app for Telegram, phone mockup with grocery list

Lee developed the app over approximately three weeks, working primarily on weekends while employed as a principal mobile architect for a Texas-based AI company. He noted that the idea had been percolating for years before he wrote the first line of code.

How CivicLoon Works

CivicLoon pulls bill text, news coverage, and committee schedules directly from the Minnesota Legislature, presenting them to users in plain-language summaries. The app currently supports 30 languages, though Lee acknowledges that translation quality varies depending on the availability of online source material in each language.

One notable technical feature is that CivicLoon runs its AI model directly on users' phones rather than sending data to remote servers. "It has no cloud reliance," Lee explained. "It only depends on itself." This design choice prioritizes both privacy and reliability, mirroring Lee's professional work developing AI chatbots that keep sensitive data off third-party servers.

Addressing AI Concerns and Limitations

While AI models can sometimes produce biased or incorrect responses, researchers note that summarization and translation are areas where current models generally excel. Daniel Schwarcz, a University of Minnesota law professor who studies AI and the legal system, explained that summarization represents a relative strength of AI technology.

"There's a lot of research that one context in which AI is pretty darn good is in summarizing text that you give it," Schwarcz said. He expressed "comparatively less worry" about bias in this application compared to others, such as recidivism-prediction algorithms that can compound historical inequities.

Lee has been candid about the app's current limitations. A feature designed to assess a bill's chance of passage remains rough, partly because AI models work from legislative text that often includes promotional language. He is working on improvements, including better bill tracking, broader translation coverage, and analysis of legislators' newsletters and public statements to provide more reliable predictions about legislative momentum.

Support for Civic Engagement Tools

Amy Perla, executive director of the League of Women Voters Minnesota, expressed support for tools that boost public participation in the political process, though she has not personally used CivicLoon. "Any way we can increase civic engagement and civic education, we think is worth exploring," Perla stated.

The app arrives as civic groups seek ways to enhance voter education and engagement, particularly during the final month of the legislative session when lawmakers determine which bills have realistic prospects for passage. While existing bill summaries on official House and Senate websites are often brief, written in legalese, and only available in English, CivicLoon aims to provide more accessible alternatives.

Schwarcz noted that the primary question is not whether the app can perform its claimed functions, but whether the public will show sufficient interest in CivicLoon compared to other products like Google's AI Overview of search results.

Pickt after-article banner — collaborative shopping lists app with family illustration

As Minnesota's legislative session progresses, tools like CivicLoon represent emerging approaches to making government more transparent and accessible to ordinary citizens who might otherwise feel disconnected from the political process.