Google is vying with Microsoft and other tech giants to bring AI capabilities to a wider audience of office professionals. The introduction of new Gemini features into Workspace marks progress toward this objective, according to analysts, though competitors are also advancing in this space with their own applications.
On Wednesday, Google unveiled a range of new functionalities that enable the Gemini AI assistant to generate and modify content within Workspace applications like Docs, Slides, and Sheets.
Julie Geller, principal research director at Info-Tech Research Group, commented that these updates signify “incremental improvements rather than revolutionary features, yet they tackle genuine workflow deficiencies. The true benefit lies in Google integrating AI assistance directly into the daily tools users rely on, such as Docs, Sheets, and Slides.”
Gemini’s “help me create” function within Docs allows it to generate a draft document based on user commands, synthesizing data from other Workspace files including existing documents, chat logs, and emails. This feature is accessible through a side panel or the newly added prompt bar at the bottom of a document. After a draft is produced, Gemini can further refine particular text sections within that document.
Gemini also offers the capability to standardize the writing style throughout a document, proving especially valuable when multiple team members contribute to its creation. Furthermore, a “match doc” feature facilitates the integration of new information into an established template.
Yulie Kwon Kim, Google’s vice president of product for Google Workspace, explained in a briefing that for numerous office professionals, the primary obstacle in initiating a project involves tedious preparatory tasks and “sifting through emails to get a preliminary draft onto the page.” She added, “Gemini now takes care of this for you, compiling information from your Drive, Gmail, and Chat to provide customized outputs tailored to your specific context. Once a draft is ready, you can then collaboratively edit it with Gemini directly within the document.”
In Sheets, Gemini can transform raw spreadsheet data into a dashboard complete with visualizations, all initiated through natural language prompts. As it executes a task, Gemini outlines a sequence of steps for user approval and poses clarifying questions when necessary.
Kim stated that Gemini and Sheets are “transitioning from a tool you simply use to a collaborative partner.” She elaborated, “Using just one prompt, Gemini gathers pertinent data from your Gmail, Chat, and Drive to generate a formatted spreadsheet in mere seconds.”
Gemini allows users to articulate their requirements in plain English, eliminating the need for intricate formulas. Furthermore, a “fill with Gemini” feature in Sheets enables rapid population of cells with pertinent data.
Google provided a company research document as an illustration. Gemini can identify the necessary information, such as HQ location, revenue, and market cap, from column headers, and subsequently search the internet to retrieve that data.
Although this tool has the potential to save office professionals a significant amount of time, the issue of hallucinations remains a persistent concern with generative AI (genAI), especially when dealing with critical business information. Google asserts that Gemini in Google Sheets has demonstrated near-human accuracy levels in benchmark evaluations, and users are provided with source links to independently verify the data.
Google says Gemini in Google Sheets has achieved near-human levels of accuracy in testing.
Amy Machado, senior research manager at IDC, commented, “Accuracy stands as a primary hurdle for genAI, and earning user trust is the ultimate challenge. Irrespective of the provider, even the most advanced tools will not gain significant widespread use without it.”
Machado noted that, with this in mind, Google is emphasizing the “context stack,” concentrating on “optimizing document AI, RAG architectures, and vector embeddings that drive these latest functionalities.
“The objective is high-precision, well-grounded retrieval and results that inspire user confidence,” Machado stated.
Gemini Capabilities in Drive and Slides
Updates have also been rolled out for Google’s Drive storage. When users submit search queries in Drive, AI Overviews presents a compilation of pertinent files along with concise summaries of their content. For more detailed inquiries, “Ask Gemini in Drive” allows for the creation of customized file repositories based on particular files, folders, or even other Workspace applications.
Mike Leone, practice director for data, analytics & AI at Omdia, described the Gemini functionalities in Google Drive as the “sleeper story” of this week’s updates. He commented, “Transforming Drive from a passive storage solution into something resembling an active knowledge-base has the potential to reshape how individuals interact with their own information.”
Leone added, “I have observed every major platform striving to tackle the challenge of ‘I know this is somewhere in my files,’ and Google’s strategy of integrating AI Overviews with Drive search strikes me as the most intuitive execution to date.”
Within Slides, users can modify individual slides using natural language prompts directed at Gemini, allowing them to adjust existing design elements and text according to their instructions. This includes the capability to generate a new slide that maintains the consistent design and formatting of the entire presentation.
Google has indicated that, in the future, users will be able to create an entire presentation from the ground up, drawing content directly from Workspace documents.
These features are being progressively rolled out this week in a beta phase and will be accessible to AI Ultra and Pro subscribers, as well as Workspace clients participating in the Gemini Alpha program.
Geller noted that these functionalities streamline workflows and offer potential time savings on recurring tasks. She explained, “Rather than toggling between applications or beginning from zero, employees now receive integrated support for drafting, data organization, and information summarization.”
However, she also pointed out “friction points” that might negate some of the potential time efficiencies. Geller stated, “The ‘Ask Gemini’ function can produce overly verbose responses without consolidation, forcing users to manually edit redundant content. Additionally, fundamental actions like pasting screenshots for contextual information do not operate as they do on other platforms, thereby restricting the AI’s overall utility.”
Leone characterized the individual features as “robust but not revolutionary, given that AI-assisted writing and slide creation have existed for some time. The truly compelling aspect is Gemini’s newfound ability to draw information from Gmail, Drive, Chat, and Calendar to construct an initial draft or compile a spreadsheet.
He elaborated, “Appropriate context is paramount in enterprise AI today, and Google possesses an inherent advantage since a substantial portion of an individual’s work context already resides within the Workspace ecosystem. I anticipate significant business interest, particularly concerning Sheets and Drive.”
Rivalry with Microsoft 365 Copilot
Google’s recent Gemini announcements closely succeed the updates to Microsoft 365 Copilot, which were revealed on Monday and included its Copilot Cowork AI agent.
Machado observed, “Regarding document applications and content generation, Google’s announcements align with Microsoft’s offerings. Microsoft, however, seems to be distinguishing itself by fostering more multi-modal and cross-application, integrated agentic experiences.”
Geller suggested that instead of directly rivaling Microsoft 365 Copilot, Google is adopting an alternative strategy. She explained, “Google isn’t merely catching up; they are pursuing a distinct path.”
Geller noted that Microsoft primarily focuses on user-centric workflow automation, whereas Google is “providing both streamlined assistants within Workspace and developer tools for crafting agents. “They’re not yet in direct competition. If these features perform consistently and save time, they will solidify user commitment to the ecosystem, a goal both Google and Microsoft share. Both companies are innovating rapidly, but speed alone will not decide the victor. The genuine distinction will stem from the design of customer feedback mechanisms, not merely the release of new features.”
Leone highlighted that while Microsoft characterized Copilot as an “agent” during its Wave 3 introduction, Google avoided using this term for Gemini in Workspace. He stated, “They present Gemini as a ‘collaborative partner,’ which indicates its current position on the AI maturity curve.”
“Based on my recent discussions with Google’s Workspace and Gemini teams, I understand their overarching vision for agentic AI is considerably more ambitious, so further developments should be anticipated,” Leone remarked. “Google arguably possesses the most robust foundational AI model and agent development frameworks currently available, and today’s features demonstrate Gemini’s excellent capability to link context throughout the Workspace ecosystem. “However, Microsoft is dictating the speed for agentic AI in the professional environment, and that is the disparity Google must address.”
