Publication voice — the personality and style that makes your outlet's writing recognisably yours — is built over years and is one of the most valuable editorial assets a newsroom possesses. Generic AI tools, left to their own defaults, produce generic prose that erodes this voice. Maintaining voice through AI-assisted content requires active style enforcement at the prompt level.
Four Voice-Preservation Techniques
1. Voice-defining system prompt: Create a detailed system prompt that captures your publication's voice with examples. Include sentence length ranges, vocabulary preferences, headline style, and tone descriptors. 2. Style guide embedding: Paste the most relevant sections of your style guide into the prompt for every draft. 3. Before/after examples: Include 3 pairs of "before (generic) / after (our voice)" examples in the system prompt. 4. Human voice-edit pass: Reserve a dedicated editing pass for voice — after accuracy and completeness are confirmed — where editors rewrite any passages that don't sound like the publication. This pass takes 10–15 minutes for most articles.