The Numbers Don’t Lie: What OpenAI’s Latest Research Reveals About How People Actually Use ChatGPT
New research from OpenAI provides unprecedented insights into AI adoption patterns that insurance executives need to understand
By James W. Moore, Founder of InsuranceIndustry.ai
When ChatGPT launched in November 2022, few could have predicted its meteoric rise. But new research from OpenAI reveals usage patterns that should fundamentally change how insurance leaders think about artificial intelligence implementation.
The Scale Is Staggering
The numbers from OpenAI’s comprehensive study are jaw-dropping. By July 2025, ChatGPT reached 700 million weekly users—roughly 10% of the world’s adult population. These users are sending 2.5 billion messages daily, or about 29,000 messages every second.
To put this in perspective, this represents the fastest global technology adoption in recorded history. For insurance executives still debating whether AI is a passing trend, these adoption rates should settle the question definitively.
The Work vs. Life Balance Surprise
Here’s where the research gets particularly interesting for our industry. Despite all the focus on workplace AI productivity, only 27% of ChatGPT usage is actually work-related as of June 2025. Even more telling, this percentage has been declining—it was 47% just one year earlier.
This trend reveals something crucial: AI’s impact extends far beyond the workplace. People are integrating these tools into their personal decision-making processes, which has profound implications for how they’ll interact with insurance products and services.
What People Actually Use ChatGPT For
The research identified three dominant use cases that account for nearly 80% of all ChatGPT interactions:
1. Practical Guidance (29% of usage)
This includes personalized advice, tutoring, and step-by-step guidance. Think of it as having a knowledgeable advisor available 24/7 for any topic.
2. Seeking Information (24% of usage)
Traditional research and fact-finding activities that compete directly with web search engines—but with the advantage of conversational, contextual responses.
3. Writing Assistance (24% of usage)
Document creation, editing, and communication support. Notably, two-thirds of writing requests involve editing existing text rather than creating new content from scratch.
The Decision Support Revolution
Perhaps the most significant finding for insurance leaders is that ChatGPT primarily provides value through decision support rather than task automation. The research shows users leverage AI for:
- Information processing and analysis
- Problem-solving guidance
- Creative thinking support
- Advisory conversations
This pattern holds remarkably consistent across different occupations and industries, suggesting a fundamental shift in how people approach complex decisions.
Professional Usage Patterns Tell a Story
When examining work-related usage specifically, writing dominates at 40% of professional interactions. This reflects AI’s unique capability to generate and refine digital outputs—something traditional search engines can’t do.
Educated professionals in high-paying occupations show the highest work adoption rates and demonstrate a clear preference for “asking” ChatGPT for advice rather than having it “do” tasks directly. Messages seeking guidance consistently receive higher satisfaction ratings than task-completion requests.
The Demographics Are Shifting
Early ChatGPT adoption skewed heavily male (around 80% in the first few months), but this gender gap has completely closed. By June 2025, slightly more users had typically feminine names than masculine ones.
Usage remains highest among younger demographics, with nearly half of all adult messages sent by users under 26. However, the research shows particularly strong growth in lower- and middle-income countries, indicating AI’s democratizing potential.
What This Means for Insurance
These findings have three critical implications for insurance industry leaders:
1. Focus on Advisory Applications
Rather than pursuing AI implementations focused purely on task automation, prioritize applications that enhance decision-making capabilities. This could include underwriting support systems, risk assessment guidance tools, and customer advisory platforms.
2. Enhance Communication Capabilities
The prevalence of writing assistance suggests significant opportunities for improving customer communications, policy explanations, and internal documentation processes. AI can help make insurance language more accessible and personalized.
3. Prepare for Informed Consumers
With 70% of AI usage being personal rather than professional, consumers are becoming increasingly comfortable with AI-powered research and decision support. They’ll likely expect more sophisticated, conversational interactions with insurance providers.
The Quality Factor
The research reveals that “asking” messages—where users seek guidance and advice—consistently receive higher satisfaction ratings than “doing” messages where users request task completion. This reinforces AI’s current strength as an advisory tool rather than a replacement for human expertise.
For insurance professionals, this suggests the most successful AI implementations will be those that augment human judgment rather than attempt to replace it entirely.
Looking Forward
OpenAI’s research provides the most comprehensive view yet of how people actually use generative AI in their daily lives. The patterns revealed—from the dominance of decision support applications to the rapid growth in personal usage—should inform how insurance leaders approach their AI strategies.
The data is clear: AI adoption is accelerating globally, users prefer advisory over automation, and the technology’s impact extends far beyond traditional workplace applications. Insurance companies that recognize and adapt to these patterns will be best positioned for the AI-driven future of our industry.
This analysis is based on “How People Use ChatGPT” by Chatterji et al., published by OpenAI in September 2025. The study analyzed over 1 million conversations using privacy-preserving automated classification methods.
About the Author: James W. Moore has over 40 years of experience in the insurance industry, working with carriers, agencies, and wholesalers. He holds a bachelor’s degree in finance with a specialization in insurance and is the founder of InsuranceIndustry.ai, a comprehensive resource covering AI applications in insurance.
AI Disclaimer: This blog post was created with assistance from artificial intelligence technology. While the content is based on factual information from the source material, readers should verify all details, pricing, and features directly with the respective AI tool providers before making business decisions. AI-generated content may not reflect the most current information, and individual results may vary. Always conduct your own research and due diligence before relying on information contained on this site.
