Google's Gemini Deep Research now has the capability to access your workspace files, marking a significant step in the integration of AI tools with personal data repositories. This feature, available to all subscribers, allows the AI to read emails, scan Google Drive files, and review Chat conversations to generate reports. However, it comes with important privacy considerations that users should be aware of before granting access.
The integration of Gemini Deep Research with the Workspace app is a notable development, joining similar capabilities offered by competitors like Anthropic's Claude and ChatGPT. This trend of AI tools accessing personal data is becoming increasingly common, raising questions about data handling and privacy.
Dave Citron, senior director of product management for Google's Gemini service, confirmed the feature's availability. The enhanced functionality enables Deep Research to pull information from Gmail, Drive (including Docs, Slides, Sheets, and PDFs), and Google Chat, along with web sources, when generating reports. This multi-step research plan, created after users enter a query, is then analysed by the AI, taking several minutes to process.
Google's announcement highlights the system's ability to mimic human research behaviour, continuously refining its analysis and initiating new searches based on discoveries. The final report includes source citations, which can be exported to Google Docs. Users can select specific sources through the Tools menu in Gemini on desktop, with mobile users receiving the feature in phases.
Privacy concerns are a significant aspect of this development. Google's spokesperson clarified that data accessed via connected apps is not used to improve generative AI. However, the privacy notice includes a critical caveat, warning users not to enter confidential information that they wouldn't want a reviewer to see or Google to use for service improvement. This directly contradicts the initial statement, raising questions about the handling of private data.
The privacy notice also advises users not to rely on Gemini's responses for professional advice, positioning Deep Research as a convenience tool rather than a trusted advisor for important decisions. This is despite its access to extensive personal and professional data, leaving users to question who benefits from this access.
Real-world performance has been mixed, with reviews ranging from glowing to sceptical. Leon Furze, an education consultant and PhD candidate, offered a sharp assessment, suggesting the tool is designed for producing lengthy, seemingly accurate reports that no one will read, focusing more on the appearance of research than actual research.
Google isn't alone in offering research AI tools, with OpenAI and Perplexity providing similar functions. The competitive landscape indicates that AI-assisted research is becoming a standard feature, with each provider having different approaches to privacy and data access.
Google's Deep Research runs on Gemini 2.5 Pro, combining Google Search with Gemini's reasoning capabilities and a one-million-token context window. The company describes it as an evolution in AI capabilities, designed to save time by tackling complex tasks. Users can access the feature by toggling the model dropdown and entering their research question.
For developers and enterprises, Gemini Deep Research with Workspace integration presents opportunities and considerations. The ability to combine internal documents, email threads, and team communications with web research could streamline workflows, but the acknowledged limitations suggest the need for oversight and validation processes rather than treating the tool as a definitive solution.
In conclusion, while this technology represents an evolution in AI capabilities, users should approach it as a productivity assistant that requires human judgment and verification, especially when handling sensitive or business-important information.
