When we attribute a statement, position or quote to a named source, that material comes from interviews, transcripts, published statements or direct engagement with documents reviewed by the reporter. AI tools should not be used to generate, extract, or summarize content that is attributed to a named source, whether direct quotation, paraphrase, or characterization of someone’s ideas.
We do not publish claims based solely on AI-generated summaries, and journalists cannot present any content as a “review” unless they have directly investigated it.
Every writer who uses AI tools in the course of reporting a story must disclose that use to their editors, and writers remain solely responsible for their content.
Images, Audio and Video
Our visual content, including a catalog of images, illustrations and videos, is produced by our editorial and art teams or obtained from photography services and wire providers. Our creative team may use AI tools in the production of some visual content, but creative direction and editorial decisions are human-driven.
We do not publish AI-generated images, audio or videos as authentic documents of actual events. We do not make changes to documentary media that alter their meaning. Standard production functions like color correction, cropping and contrast adjustment are fine.
When synthetic media is used in the context of reporting on AI, it will be clearly identified as AI-generated, with that disclosure kept as close to the content as possible.
Accountability cannot be compromised
Anyone who uses AI tools in our editorial workflow is responsible for the accuracy and integrity of the resulting work. This responsibility cannot be transferred to colleagues, editors or the tool itself. More broadly, maintaining the standards set out in this policy is a shared responsibility across our editorial operations.
These standards have governed our editorial work since AI tooling became available. When violations occur we take action. We are publishing this reader-facing version because our readers deserve to see the rules we follow, not just trust that they exist.
This policy was last updated on April 22, 2026.
<a href