Google is testing how much companies will pay for AI already inside their files

Google's March 10 beta isn't subtle about its ambitions: Gemini is now embedded across Workspace, turning Docs, Sheets, Slides and Drive into a unified generative playground. The flagship addition, "Help me create" in Docs, scans your Drive, Gmail, Chat and web history to produce structured drafts that mirror the formatting of reference documents you already own. Google notes that over a third of new Docs start as copies; this feature automates the mimicry, matching headers, spacing and tone without the manual slog.

Sheets graduates from formula helper to full generator: describe a budget tracker or sales forecast in plain English, and the spreadsheet materializes with labeled columns, conditional formatting and working calculations. Slides follows the same trajectory, promising complete presentations from single prompts—though Google has been cagey about timing, suggesting a staggered rollout that leaves power users waiting. Drive's "Ask Gemini" finally lets you interrogate your own files with natural language, a feature that feels embarrassingly overdue for anyone managing thousands of documents across shared workspaces.

The technical architecture is telling. Rather than training fresh models, Google is leveraging its existing moat: decades of user files, organizational patterns and cross-app behavior. The beta represents a repurposing of accumulated data assets through a conversational interface, not a fundamental breakthrough in capability. This is integration as innovation, the kind of move only a platform with Google's reach can execute at scale.

The barriers to entry are deliberately steep. Access requires Google AI Ultra, Pro or Gemini Alpha business subscriptions—the most expensive tiers—limiting the beta to enterprises already deep in Google's ecosystem. English-only support further narrows the field, a reminder that even trillion-dollar companies still treat multilingual deployment as an afterthought.

Microsoft's 365 Copilot has already staked similar ground, offering drafting and data analysis across Office apps. Google's differentiation lies in cross-app context: a prompt in Docs can pull from your Gmail threads and Drive spreadsheets simultaneously, creating a coherence that siloed tools struggle to match. Whether this translates to genuinely better output or merely more confident-sounding mediocrity remains the open question.

The historical pattern invites skepticism. Previous Workspace AI features have lingered in beta limbo, promising transformation while delivering incremental convenience. The "Help me create" feature risks the same fate: impressive in demo, uneven in production, particularly for users whose reference documents are themselves inconsistent or poorly structured.

For administrators, the calculation is straightforward. The premium pricing buys early access and potential productivity gains, but also commits organizations more deeply to Google's data collection apparatus. The irony is thick: you're paying extra for the privilege of having your own files analyzed to produce work that still requires human verification. The beta succeeds if it genuinely reduces cognitive overhead; it fails if it becomes another layer of automation that demands its own form of babysitting. Google's bet is that the former outweighs the latter enough to justify the subscription premium before competitors close the gap.