5.3. AI in Research: Expanding Inquiry and Insight
Jennifer Martin
Artificial intelligence is increasingly woven into the research enterprise. From medical breakthroughs to humanities scholarship, AI accelerates discovery and expands what is possible. Importantly, AI does not replace human inquiry—it reshapes the tools, pace, and questions we bring to knowledge-making.
AI as a Research Partner
Key uses of AI in research include:
- Data Analysis at Scale — processing large datasets in seconds, spotting patterns or anomalies humans might miss.
- Literature Review Support — summarizing and synthesizing hundreds or thousands of papers to map out themes or research gaps.
- Visualization — generating charts, graphs, visual models, sometimes even interactive dashboards to reveal insights.
- Hypothesis Generation — suggesting new directions for inquiry based on trends in existing research.
- Drafting & Communicating Findings — helping produce summaries, translating text, assisting in editing for clarity or style.
Here are some example prompts for using AI as a research partner:
- Prompt – Literature Review Organization
- “Summarize and cluster the following abstracts into thematic groups related to resilience in graduate health professions education. Provide a one-sentence description for each theme.”
- Prompt – Data Visualization Support
- “Explain how to visualize qualitative interview data on student resilience using thematic mapping. Suggest software options and describe a simple figure format for publication.”
- Prompt – Clarity and Tone Revision
- “Revise the following research abstract for clarity and conciseness while maintaining an academic tone appropriate for The American Journal of Occupational Therapy.”
Ethics, Limits, & Responsibilities
AI tools offer remarkable potential in research but also require careful ethical oversight. The following principles outline key responsibilities for scholarly integrity:
- Accuracy & Verification — AI can produce plausible but incorrect or misleading claims. Researchers must verify, cross-check, and avoid assuming correctness.
- Bias and Representation — Models reflect the data they are trained on. Underrepresented voices and perspectives may be diminished or misrepresented.
- Transparency and Authorship — Clearly describe how AI contributed to text, data, or analysis. Disclose its use in publications and methods sections, and maintain clear human authorship.
- Authorship & Credit — how should AI contributions be acknowledged? This is still unsettled in many fields.
- Licensing, Privacy, and Permissions — Ensure compliance with copyright, licensing, and privacy standards, particularly when using human subject data or sensitive content.
- Reproducibility — Document the tools, versions, and parameters used so that AI-supported findings can be replicated.
Expanding the Research Conversation
Beyond efficiency, AI is changing how we think about research itself. These shifts bring both opportunity and responsibility:
- AI can accelerate discovery, but speed must not replace rigor. The obligation to confirm findings, replicate analyses, and evaluate accuracy remains central to scientific integrity. Because AI systems learn from historical data, biases in representation persist, reinforcing gaps in whose voices and experiences are reflected in research outputs.
- As researchers experiment with AI in writing and analysis, questions of authorship and credit are evolving. The most ethical approach is transparent acknowledgement of where AI was used and where interpretation remained human. Likewise, privacy and data governance must remain priorities, since not all accessible data are ethically or legally available for processing.
- Finally, this changing landscape calls for new research literacies. Skills such as designing effective prompts, critically assessing AI outputs, and comparing performance across tools are becoming essential competencies. Like statistical literacy in the past, these emerging skills will define credible and ethical scholarship in the AI era.
📖 Analogy: AI as a Microscope for Knowledge (click to expand)
Just as the invention of the microscope revealed entire worlds invisible to the naked eye, AI enables researchers to see patterns, correlations, and possibilities that human cognition alone might miss. Yet, like the microscope, it is a tool; what we see depends on how we use it, and interpretation remains a human responsibility.
Video Resource
Watch this video that shows how academics are using AI ethically and productively in their research work:
Interactive Activity
Decide which of these research tasks are Ethical Use, Use with Caution, or Avoid. Drag each into the category that best reflects responsible research practice.
Applied Research Challenges and Practice Suggestions
Here are more nuanced scenarios from some research domains; think about how you might apply AI tools in each case.
- A researcher uses AI to generate a first draft of the background section. Should they disclose it, and how should they revise it?
- Collecting data via AI scraping from websites—what legal and ethical checks are needed?
- Translating interviews collected in local languages with an AI tool—how to ensure accuracy and preserve participants’ voice?
- Using AI to help visualize data that has sensitive metadata—how to anonymize properly and respect privacy?
📚 Weekly Reflection Journal
Looking Ahead
Next, in 5.4 AI in Productivity and Personal Life, we will explore how AI tools can enhance work-life balance, personal efficiency, and daily living, while preserving ethics and meaning.
