AI-generated images and videos are becoming increasingly convincing, prompting journalism to adapt by implementing verification to ensure authenticity.

There have been real-world cases of AI misuse in real estate visuals and marketing photos, with some corrections made post-publication or pre-publication checks preventing misuse.

Freelancers and internal teams are being trained on AI policies, with clear guidelines for editors, photographers, and agents to prevent misleading visuals.

The Globe has updated policies to ban generative AI for creating, editing, or enhancing visuals, require disclosure when AI is used, and prohibit large-scale edits that change meaning.

The piece highlights the media’s duty to maintain accuracy and urges readers to verify images before sharing, while pointing to resources for detecting AI content.

The Globe’s verification process now uses multi-step checks: corroborating with other location-specific images, confirming subject whereabouts, and employing AI-detection tools like Hive AI Detector.

Examples of AI risk include a July image of Mitch McConnell circulating as a life-support hoax and AI-edited real estate photos used to mislead buyers or boost listings.