Edit locality is the ability of an image model or editor to change one targeted region while leaving the rest of the picture stable. Good locality means a user can replace a logo, fix a hand, or adjust a background without causing unwanted drift in faces, objects, lighting, or layout.
In cyber security, edit locality matters because it affects trust, moderation, and forensic review. Strong locality can help defenders create controlled training data, redact sensitive details, or test detection systems with precise synthetic changes. The same capability can also support abuse: attackers can alter small visual details in a convincing document, screenshot, or product image while preserving the surrounding context. Security teams should look for region-level change analysis, provenance checks, and human review when images are used as evidence or shared in workflows.