Adobe introduced a major update to Acrobat’s artificial intelligence assistant, which it introduced in February. This update included AI-generated summaries and a chatbot that can answer questions about the content of large documents. Adobe will add new capabilities to this chatbot, giving it the ability to analyze and query multiple documents starting tomorrow.
Adobe adds multiple document analysis and rendering features to Acrobat’s AI assistant
Adobe is also adding an AI-powered image creator to its Acrobat software. This feature, supported by the Firefly model, will allow users to remove backgrounds or make minor changes within existing PDF documents.
Adobe states that these AI models are “safe for work” and will not compromise user privacy or create images that will negatively impact a brand’s image.
The multi-document artificial intelligence analysis feature added to Acrobat stands out as the most interesting part of this update. Previously, the assistant focused only on individual documents and provided users with answers linked to resources within the documents.
Adobe will expand this feature, allowing artificial intelligence analysis on a group of documents. Users will be able to drag and drop non-PDF file types, such as Word and PowerPoint documents, into the chatbot interface. From there, they can ask questions in the usual way or have the AI identify trends across multiple documents.
While other chatbots, such as Gemini and ChatGPT, allow users to upload documents for analysis, Adobe believes users of document editing software will see the value of an integrated solution.
Update from Microsoft that makes file sharing easier between PC and Android
Microsoft simplifies file sharing between PC and Android with the update it offers to the beta channel of Windows 11.
This offer from Acrobat is also cheaper. The AI assistant costs $5 per month, less than the $20 you’ll need to pay for ChatGPT Plus or Gemini Advanced. Additionally, Adobe will offer this feature free of charge for all users between 18-28 June 2024.
Confidence in Adobe’s AI features has fluctuated in recent weeks. The company faced user backlash after admitting that it used a small sample of Midjourney images to train its Firefly model.
It later backtracked on some terms of service clauses that potentially granted the right to analyze user content. The company clarified that it will not access user content to train AI models and will prohibit third-party providers from using that data.
With this update, Adobe continues its stance on protecting user privacy. The documents will be uploaded to the cloud, but the company will not use the data to train its models. Similar to Apple’s ChatGPT integration, third-party LLM providers will also be prohibited from using the data.