How the Bulletin uses AI
The Bulletin is experimenting with how artificial intelligence (AI) can improve access to our journalism. Our use of AI is governed by strict internal guidelines prioritizing transparency, accuracy, and privacy.
We currently use generative AI only for audio versions of articles. We do not use AI to generate news content or media that could challenge the editorial meaning of news content and/or mislead audiences.
We will update this page as our use of AI evolves.
Articles read by an automated voice
The Bulletin of the Atomic Scientists offers an audio version for most newly published articles to enhance accessibility and provide more ways to engage with our journalism.
These audio recordings are created using artificial intelligence (AI) text-to-speech technology, not by a human narrator. We hope this feature improves the Bulletin’s inclusivity and usefulness for readers with different preferences and needs. All articles are still researched, written, and edited by our editorial team before being processed by the automated reader.
Because the voice is generated by AI, you may occasionally notice errors in pronunciation, tone, voice, or sentiment that would differ from a human narrator, and the audio may on rare occasions also skip words. We are committed to transparency and welcome your feedback to help us improve this service. If you encounter a significant error or have suggestions, please contact us.