Adobe has announced plans to introduce a new generative AI (genAI) assistant in Acrobat, aimed at simplifying the comprehension of lengthy PDF documents. The AI Assistant, currently in private beta, will eventually be integrated into the paid Acrobat desktop and web applications, as well as the free Reader tool. Users can expect to require an add-on subscription for access, although specific pricing details have not been disclosed yet.
According to Rosanna Jimenez, a research analyst at S&P Global Market Intelligence, the AI Assistant in Acrobat will be particularly beneficial for teams and individuals dealing with text-heavy documents in fields such as legal, research, editorial work, and archival institutions. While there are existing open-source tools that offer similar functionalities for isolated tasks, having a standardized AI tool within Acrobat can enhance cohesion and accuracy for teams and organizations.
Accessible through a sidebar within the application, the Acrobat AI Assistant offers document overviews with concise summaries of content across different sections, including text, graphs, and infographics. Users can click on these summaries to navigate to specific sections for more detailed information. Additionally, the AI assistant allows users to ask questions related to the document, providing relevant suggestions and answers with citations linking back to the document content.
Notably, the AI assistant’s responses are limited to information within the document itself and do not incorporate external web-based context. Users can also format summarized information for easy sharing via email or other platforms. Adobe assures that documents analyzed by the AI Assistant are not stored or utilized for training its language models.
Future enhancements for the Acrobat AI Assistant include broader document type summarization, encompassing Microsoft Word and PowerPoint files. Adobe also plans to expand the AI assistant’s capabilities to support authoring and editing tasks within Acrobat, such as generating initial drafts or adjusting voice and tone. Furthermore, integration with Adobe’s Firefly generative AI tool will enable the addition of images to text documents.
Ritu Jyoti, IDC’s group vice president for Worldwide Artificial Intelligence and Automation services, anticipates a significant demand for the Acrobat AI Assistant, particularly in knowledge discovery applications. While concerns about hallucinations exist with any generative AI service, Jyoti highlights Adobe’s approach of grounding responses in document content and providing verifiable results with proper attribution to mitigate such risks.
