The rise of generative AI models like ChatGPT has radically transformed the landscape of information access and communication. In the past, users seeking information relied heavily on traditional search engines, digital libraries, or human experts. Today, conversational AI has become a ubiquitous tool used across varied contexts—home, education, work, leisure—for a wide range of purposes.
In October 2024, ChatGPT Search was introduced. It allows ChatGPT to search the web in an attempt to make more accurate and up-to-date responses.
This study explores how ChatGPT meets the information needs of users through a qualitative content analysis of 205 user vignettes. The goal is not merely to understand what questions users are asking, but to uncover the deeper practices and behaviors that arise when engaging with an AI-powered assistant. This article expands upon the original findings by examining the six identified information practices (Writing, Deciding, Identifying, Ideating, Talking, and Critiquing) in greater detail, while offering a theoretical perspective on how "information needs" are being redefined in an age of human-AI collaboration.
Historically, information retrieval has been conceptualized as a linear process: a person has a question, they seek an answer from a source (a book, a website, a database), and they evaluate the results. With conversational agents like ChatGPT, this process becomes fluid and dialogic. The user may begin with a vague idea, refine their question through interaction, and receive not only factual information but also contextual suggestions, critiques, and empathetic responses.
This shift has critical implications for understanding "information needs." Instead of static, well-defined queries, needs are often emergent and shaped by dialogue. In many cases, users don't fully know what they want until they begin a conversation.
The study revealed that ChatGPT is utilized across multiple domains of life:
Home & Family: Parenting advice, relationship counseling, recipes, home repairs.
Work: Writing reports, automating emails, analyzing data, drafting business proposals.
Education & Learning: Explaining difficult concepts, summarizing readings, preparing for exams.
Leisure: Creative storytelling, gaming strategies, travel planning.
Health & Wellness: Diet tips, mental health support (non-clinical), habit tracking.
Each domain reflects a blend of emotional, cognitive, and pragmatic needs. Importantly, users often switch between these roles fluidly—for instance, asking for emotional advice one moment, then requesting a Python script the next.
Through content analysis, six dominant categories of user interaction emerged. Each represents a different modality of how users interact with ChatGPT and how the AI supports their activities.
This is perhaps the most common use case. ChatGPT helps users write or edit a wide range of texts: emails, academic essays, blog posts, CVs, cover letters, code documentation, even poetry.
Writing assistance includes:
Drafting content from scratch
Polishing grammar and tone
Adapting content for different audiences or formats
Translating between languages
Converting ideas into structured formats (e.g., outlines)
Example:
“I need to write a letter to my landlord about a plumbing issue. Can you help me phrase it professionally?”
The AI becomes a co-author, proofreader, and stylistic advisor.
ChatGPT often supports decision-making processes where users seek comparisons, pros and cons, or strategic recommendations.
These decisions include:
Choosing between job offers
Picking a vacation destination
Selecting a college major
Evaluating software tools
Considering lifestyle changes
Example:
“Should I switch from Notion to Obsidian for my personal knowledge management? What are the trade-offs?”
Here, ChatGPT serves as a decision support system, helping users reason through options by providing structured arguments.
This practice involves categorizing, classifying, or diagnosing something. Users often want to know what something is or what category it belongs to.
Examples include:
Identifying logical fallacies in arguments
Classifying types of learning disabilities
Spotting themes in a literary text
Recognizing patterns in code or data
Example:
“What kind of anxiety disorder does this list of symptoms suggest?”
ChatGPT's role here is analytic and comparative, often drawing from domain knowledge.
Users use ChatGPT as a brainstorming partner for creative or strategic generation of ideas.
Common ideation tasks:
Coming up with business names
Generating social media captions
Creating plot ideas or characters
Planning fundraising campaigns
Designing learning activities
Example:
“Give me 10 creative TikTok video ideas for promoting my bakery.”
Ideation blends creativity with domain understanding and audience awareness.
Some users interact with ChatGPT not to get information, but to talk. They want to reflect, vent, role-play, or simulate a conversation.
This includes:
Practicing interviews or presentations
Simulating dialogue with a historical figure
Engaging in emotional self-reflection
Exploring philosophical questions
Example:
“Pretend you are Socrates. Let’s talk about justice.”
Here, the AI is more like a conversation partner or philosophical mirror than a source of knowledge.
Finally, many users come to ChatGPT with work-in-progress and ask for critique.
Critiquing includes:
Reviewing essays or code
Evaluating logic or clarity
Suggesting improvements
Giving feedback on tone or persuasiveness
Example:
“Here's my draft cover letter. Can you suggest ways to improve the opening paragraph?”
This transforms ChatGPT into an editorial partner or even a digital mentor.
Traditional models of information behavior, such as Wilson’s model or Kuhlthau’s ISP model, tend to frame information seeking as either a problem-solving activity or a sense-making journey. In the AI age, we must broaden this perspective.
Information need is no longer just about getting the right answer.
It is about coping skillfully with the challenges of daily life—whether cognitive, emotional, social, or professional.
Users approach ChatGPT not just to know, but to:
Act (e.g., write a resume)
Reflect (e.g., talk about emotions)
Express (e.g., write poetry)
Decide (e.g., change careers)
Create (e.g., launch a business)
In this view, "information need" becomes intertwined with the human need to perform, decide, grow, and relate.
Given ChatGPT’s conversational tone, users often attribute authority to its responses, even when the system explicitly warns of potential inaccuracies. This raises questions about digital literacy and the ethical design of AI systems.
Many users turn to ChatGPT for emotional support, though the system is not a substitute for professional mental health care. How should generative AI respond to vulnerable users in distress?
Used skillfully, ChatGPT can empower individuals—especially those with limited access to professional expertise. For example, a student with no access to tutors can still get help with calculus. An immigrant can get assistance writing official letters in a second language.
Rather than replacing human knowledge workers, ChatGPT is most effective as a collaborative tool. It augments rather than replaces human capabilities.
This study opens numerous avenues for further exploration:
Longitudinal studies on how individuals integrate ChatGPT into daily life over time.
Comparative studies on how ChatGPT is used differently across cultures, age groups, or professions.
Ethnographic studies exploring how trust in AI is formed and negotiated.
Interface design research to explore how UI/UX affects information practices with AI.
ChatGPT is not just a Q&A tool; it is a flexible, dialogic, and expressive platform that supports a rich array of human activities. It meets users in diverse life contexts—home, work, school, leisure—and adapts to their evolving needs. By supporting writing, decision-making, identification, ideation, talking, and critique, ChatGPT redefines what it means to seek information.
The emerging insight is clear: Information need in the AI age is not only about finding facts but about navigating life itself. Understanding, action, and reflection now go hand in hand. As AI continues to evolve, so too must our frameworks for understanding how people seek, use, and experience information.
