on this page

USBills.ai

software

Open-source platform using AI and NLP to make US federal legislation accessible through plain language summaries, ELI5 explanations, and readability metrics

period: 2025-present
team: ALEA Institute
tech:
Legal InformaticsNatural Language Processing
══════════════════════════════════════════════════════════════════

An innovative platform that democratizes access to US federal legislation by combining traditional NLP with large language models to provide clear, accessible analysis of congressional bills.

Mission

USBills.ai addresses a critical civic challenge: federal bills often consist of hundreds of pages of dense legal language, making them inaccessible to most citizens. This platform transforms complex legislation into understandable insights.

Core Features

AI-Powered Analysis

  • Plain Language Summaries: Clear, concise explanations of bill content
  • ELI5 Explanations: β€œExplain Like I’m 5” versions for maximum accessibility
  • Readability Metrics: Automated Readability Index and complexity scores
  • Financial Analysis: Impact assessments and cost estimates

Technical Analysis

  • Tokenization: Breaking down text structure
  • Syntactic Parsing: Understanding grammatical relationships
  • Named Entity Recognition: Identifying key actors and organizations
  • Parts of Speech Distribution: Linguistic complexity analysis
  • Token Entropy: Information density metrics

Data Collection

  • Daily Updates: Automatic retrieval from govinfo.gov API
  • Complete Coverage: All bills from the current Congress
  • Historical Data: Analysis of past legislation
  • Version Tracking: Changes and amendments over time

Technical Architecture

Processing Pipeline

  1. Data Ingestion: Daily pulls from Government Publishing Office
  2. NLP Processing: spaCy-based traditional analysis
  3. LLM Enhancement: AI-generated summaries and explanations
  4. Quality Control: Validation and accuracy checks
  5. API Distribution: Open access to all annotations

Technology Stack

  • XSLT (78%): Document transformation
  • Python (15.5%): Analysis and processing
  • HTML (6%): Web interface
  • spaCy: Natural language processing
  • LLMs: Advanced text generation

Metrics Tracked

For each bill, the platform provides:

  • Section and subsection counts
  • Word and token statistics
  • Readability scores
  • Complexity indices
  • Key entity extraction
  • Topic classification

Example Output

H.RES. 1603 Analysis

ELI5: β€œThe House wants to keep track of and share what they find out about bad things someone might have done when they was part of the government.”

Summary: Detailed explanation of ethics committee requirements, document preservation, and public disclosure procedures.

Keywords: committee on ethics, misconduct investigation, congressional accountability

Open API

USBills.ai provides a free, open API for:

  • Bill metadata and full text
  • AI-generated summaries
  • Readability metrics
  • Historical comparisons
  • Bulk data export

Ethical Considerations

  • Transparency: All AI-generated content is clearly labeled
  • Accuracy: Supplements, not replaces, human review
  • Source Attribution: Links to official government sources
  • Non-partisan: Objective analysis without political bias

Impact

The platform serves:

  • Citizens: Understanding legislation affecting their lives
  • Journalists: Quick analysis for reporting
  • Researchers: Data for academic studies
  • Advocates: Tracking relevant bills
  • Educators: Teaching civic engagement

Future Development

Planned enhancements include:

  • State legislation coverage
  • Amendment tracking
  • Voting prediction models
  • Multi-language support
  • Mobile applications

Data License

All data and analysis are released under Creative Commons Attribution 4.0 International License, ensuring maximum accessibility and reuse.

USBills.ai represents ALEA Institute’s commitment to using AI for public good, making democratic processes more transparent and accessible to all citizens.

on this page