The AI Observatory analyzed user-consented conversations to reveal how AI use shifts across models and time.

Stanford researchers launched the AI Observatory, a public platform that aggregates and analyzes real AI conversations from seven consent-based datasets, to provide independent usage data for researchers and policymakers.

Stanford PhD candidate Anka Reuel co-leads the project. The project emerged because companies like Anthropic and OpenAI release selective usage reports with no independent verification. The platform aims to give researchers and policymakers a clearer picture.

For builders and operators, the data reveals significant gaps in official usage reports, especially around health, relationships, and sexual content, which directly impacts content moderation and safety design.

The AI Observatory plans to expand its datasets over time, and researchers expect the independent data to shape policy debates about AI benefits and risks as the platform grows. Anthropic and OpenAI will face pressure to validate their internal usage numbers.

What matters

• Stanford researchers launched the AI Observatory using seven public conversation datasets.
• Company reports filter out nearly half of conversations, hiding sensitive AI use.
• Watch for the AI Observatory to add more datasets and reveal shifting usage trends.

Why it matters

Watch for the AI Observatory to add more datasets and reveal shifting usage trends.

This GenAI News article was prepared in original wording using reporting and materials published by MIT Technology Review AI. Source reference: https://www.technologyreview.com/2026/08/18/1142226/how-people-use-ai/.

Drafted by the GenAI News review pipeline.