Adobe announces controls to refine image and PDF edits inside ChatGPT

Image selections and adjustment controls

Adobe has announced interactive editing features for its plugin in ChatGPT, bringing selected image adjustments and document changes into side panels beside the conversation. Adobe says people working on images can identify an area, request a change through ChatGPT and then refine the result themselves. The announcement gives users a way to combine written requests with direct adjustments within the same workspace.

For image work, Adobe describes selecting an object or painting over an area with a brush before adjusting properties such as hue, saturation and contrast. The controls presented depend on the requested edit, according to Adobe. A request to recolor an object, for example, brings up color adjustments for that selection, giving the user a defined area to refine.

Conversation context and PDF edits

Adobe says adjustments made in the editor become context for the ongoing chat, while requests in the conversation determine the editing controls presented. For someone making successive changes to an image, the stated consequence is continuity: they can move between describing an adjustment and making it directly without uploading the content again or restating the work already done.

PDF editing uses Acrobat tools within the plugin, according to Adobe. Users can change document text or mark passages in the side panel, and a selection can guide a subsequent request. Adobe illustrates this with a freelance contract: a user selects payment language and asks for related payment or invoicing sections to be highlighted. Those markings bring terms together for the user to examine for consistency.

A rollout beginning on web and app

Adobe says worldwide distribution starts on the announcement day across ChatGPT’s web and app interfaces. That describes the beginning of deployment; it does not establish that every user already has the controls. The availability claim therefore supports a global rollout in progress, with the new image and document interactions presented as features of the Adobe plugin in ChatGPT.