Anthropic, a prominent AI startup, has announced two significant updates to its AI assistant Claude: a research tool and integration with Google Workspace. This move positions Claude as a more comprehensive workplace AI assistant, directly competing with Microsoft's Copilot and similar tools.

The Research feature, available in early beta for Max, Team, and Enterprise plans in the US, Japan, and Brazil, transforms how Claude finds and reasons over information. Unlike traditional AI search capabilities, Claude now operates "agentically," conducting multiple searches that build on each other to provide more accurate results. The feature includes citations, allowing users to verify the information themselves.

Anthropic's product executive, Scott White, explained that the company aims to achieve a "sweet spot" of one to five minutes per query for the Research feature. This enables users to work with Claude directly to solve problems without disrupting their workflow. White also hinted that the Research feature would soon be available to users on the $20-a-month Pro plan.

In addition to the Research feature, Anthropic has introduced Google Workspace integration, allowing Claude to connect directly to users' Gmail, Google Calendar, and Documents. This integration enables Claude to pull together meeting notes, identify action items from email threads, and search relevant documents for additional context, making it a more effective workplace assistant.

The Google Workspace integration is available in beta to all paid users, although Team and Enterprise plan administrators must enable Google Workspace access company-wide before individual users can connect their accounts. For Enterprise customers, Anthropic is also rolling out "Google Drive cataloging," which uses retrieval augmented generation (RAG) techniques to improve document search.

Two significant concerns surrounding the use of AI assistants like Claude are hallucinations and privacy. Hallucinations occur when the AI system fails to connect the dots between complex inputs, potentially providing inaccurate information. Anthropic encourages users to verify information through citations to mitigate this risk. Regarding privacy, Anthropic's Scott White emphasized the importance of user-level authentication, ensuring that Claude only accesses authorized documents based on the user's login.

Anthropic's updates come on the heels of similar announcements from OpenAI and Google, which launched their own research tools, dubbed Deep Research. The AI landscape is rapidly evolving, with companies like Anthropic, OpenAI, and Google pushing the boundaries of what is possible with AI assistants.

Anthropic's press release hints at further updates in the coming weeks, promising to expand "both the range of context available and the depth of the reporting" for Claude. As the AI assistant market continues to grow, it will be interesting to see how Anthropic's updates impact the competitive landscape and shape the future of workplace AI assistants.