Editorial
How and why we use artificial intelligence in our editorial process.
Last updated: 2026-04-25
Brick & Bit uses language models (LLMs) — currently DeepSeek — as an editorial tool. Use is limited to: (1) Spanish↔English translation; (2) structuring content into sections, subheadings, and lists; (3) rewriting to avoid verbatim copying of original sources; (4) generating excerpts and key takeaways.
AI does not invent figures, generate quotes attributed to people, produce its own financial predictions, recommend investments, or approve content for publication. Every decision about what to publish and under what criteria is made by the human editorial team.
Each AI-generated article passes automated validations (length, source, absence of promotional language). Additionally, we apply manual review on a random 10% sample of published articles to detect model drift and adjust prompts.
Every article includes a visible notice — "Content produced with AI assistance from [source]" — alongside the byline. This lets the reader contextualise the origin and apply their own judgement.
If you have questions, concerns, or feedback about how we apply AI in our process, email editorial@brickandbits.com. We publish any significant change to this policy with a visible revision date.