Blog

Generative AI is everywhere. But do we really understand where and how it’s being used?

At Quanam, Marketing freed up 2,300 hours, Legal accelerated contract reviews, and Finance now produces executive reports in minutes. Processes are genuinely faster. Yet data from 37.5 million conversations reveals another dimension: late at night, people aren’t asking about Excel, they’re asking about health, relationships, and philosophical questions.

Productivity: a promise delivered

As highlighted at EXCELSA 25, “processes that used to take weeks are now resolved in hours or minutes”. Marketing reduced content production cycles from eight months to three. Legal teams use AI to review and improve legal documents, compare versions, and generate summaries. Finance turns complex data into clear executive narratives, interprets tax regulations, and produces compliance checklists.

Is this specific to Quanam? No. In November 2025, Anthropic analyzed 100,000 real conversations and identified the same pattern. The study estimated how long professionals would need to complete, without AI, the tasks people were performing with Claude, then compared both figures to calculate actual time savings. Legal and Management led in task complexity (approximately 1.8 and 2.0 hours without AI, respectively), achieving time savings of 80.7% and 85.1%. They were followed by Education (1.7 hours, 84.5% savings) and Arts/Media (1.6 hours, 78.6%). These sectors align closely with Claude’s strengths in complex knowledge work: reading, analyzing, and synthesizing information at speed.

The study itself acknowledges limitations. It does not capture time spent validating outputs outside the chat, refining results, or coordinating with teams. The estimates also lack context regarding user expertise or specific workflows. Still, it provides a useful quantification of AI’s impact in high-value professional domains.

For those interested in going deeper, Anthropic publishes insightful research on anthropic.com/research.

While this piece focuses on business and operational areas, a previous article explored how generative AI is also transforming software and product development at Quanam, where the conversation with AI becomes the core of the build process.

The unexpected: beyond the office

We know AI isn’t used only for work or productivity. It’s also deeply personal.

In December 2025, Microsoft AI published its Copilot Usage Report 2025, analyzing 37.5 million conversations between January and September 2025. The report examines when and why people turn to AI beyond the workplace.

Health consistently ranks as the leading topic on mobile devices, regardless of day, month, or time. People aren’t just seeking medical information. They turn to AI for advice on relationships and life decisions. The report also highlights striking patterns: a sharp spike in conversations about relationships on February 14 (Valentine’s Day), and a dramatic increase in philosophical topics during the early hours of the morning.

The use of generative AI for personal and health-related matters raises a serious flag about the dependency we are developing on these technologies, and what that means for privacy, autonomy, and human well-being.

Kenji Nakasone
Lead AI Innovation Engineer at Quanam