Adobe Inc. today introduced new artificial intelligence features for Acrobat, its PDF editing application, that will enable users to generate images and analyze large document collections.

The update is rolling about two months after the company equipped Acrobat with an embedded AI assistant. The chatbot can answer user questions about a document and generate summaries. According to Adobe, the AI upgrade detailed today significantly expands upon those capabilities.

Acrobat can now answer questions not only about PDF text but also Word documents, PowerPoint presentations and other files. Moreover, it’s capable of analyzing such files in bulk. Users can drag and drop a collection of documents stored in different formats into Acrobat and then ask natural language questions about the information inside.

The built-in AI generates prompt suggestions to help users sift through documents faster. The chatbot includes citations in each prompt response to help users verify the accuracy of the retrieved information. The citations double as shortcuts for navigating to the specific sections of a document that the AI used to generate its answer.

The feature is rolling out alongside two other natural language processing capabilities. The first addition promises to automate the task of summarizing meeting transcripts. The other new feature allows workers to turn information they extract from documents into a specific format, such as a marketing newsletter, using natural language prompts.

A third set of AI capabilities included in the update will enable users to generate images for documents. Additionally, they promise to ease the task of editing a PDF file’s existing visual assets.

According to Adobe, the feature bundle includes an AI tool that can automatically remove the background of the objects in an image. A second feature addition, dubbed Generative Fill, allows users to add new objects by inputting a brief prompt describing what they wish to draw. Today’s Acrobat update likewise promises to speed up simpler tasks such as cropping and resizing images.

“With the new Firefly and AI Assistant capabilities in Acrobat, anyone can transform the information in their digital documents into actionable insights and visually compelling content – quickly and easily,” said Abhigyan Modi, senior vice president of Adobe’s Document Cloud business.

Acrobat’s new graphic design features are powered by an AI model dubbed Firefly Image 3 that Adobe debuted in April. Compared with the company’s earlier neural networks, it can generate higher-quality images and supports a broader variety of styles. The company says it’s also more adept at understanding lengthy prompts.

Image: Adobe