Baltimore Banner is testing generative‑AI tools to streamline story creation, automate transcription, and engage readers through a custom chatbot. The pilot aims to boost investigative capacity while safeguarding editorial integrity, with each tool undergoing rigorous bias and hallucination checks before any public use.
AI Pilot Projects Overview
Draft Generation for Investigative Stories
The newsroom uses large‑language models to produce first‑draft outlines for local‑government investigations, allowing reporters to focus on deep research and analysis.
Automated Transcription and Tagging
AI‑driven transcription quickly converts audio interviews into text, while keyword‑tagging algorithms organize content for faster editorial review.
Reader‑Facing Chatbot for City Initiatives
A custom chatbot answers community questions about city‑wide cycling programs, delivering instant, accurate information and freeing staff to cover more complex beats.
Ethical Guidelines and Transparency
Every AI tool undergoes a three‑step vetting process:
- Bias mitigation – systematic reviews to detect and correct skewed outputs.
- Hallucination checks – verification against trusted data sources before publication.
- Clear disclosure – readers are informed when AI assistance is used in story production.
Impact on Reporting Workflow
By handling routine tasks, AI frees reporters to pursue in‑depth investigations, enrich feature pieces, and engage more directly with the community. The newsroom reports faster turnaround times and higher staff satisfaction during the pilot phase.
Future Plans and Evaluation
After the pilot, Baltimore Banner will publish a comprehensive review covering performance metrics, audience feedback, and any policy refinements. Successful tools may become permanent fixtures, further enhancing the paper’s commitment to innovative yet trustworthy journalism.
