Responsible Conduct of Research in the Age of AI
By Dr. Zubair Khalid, DVM, MS, PhD ·
Quick Answer
- Update RCR training by adding a dedicated module on AI, big data, and gene editing ethics that maps each emerging technology to existing responsible conduct domains.
- Use the NIH Data Management and Sharing Policy as a concrete anchor for teaching data stewardship expectations in the AI era.
- A key limitation is that institutional policies vary, so instructors must adapt the module outline to their own organization's requirements and update it as federal guidance evolves.
Why RCR Training Must Address Emerging Technologies
Research integrity training has traditionally focused on stable principles: accurate recordkeeping, honest reporting, proper authorship, and transparent peer review. The Committee on Publication Ethics Core Practices describe these as foundational expectations for researchers and journals. However, the tools researchers now use daily have changed the practical meaning of those principles.
Artificial intelligence tools can draft text, analyze images, and generate code. Big data platforms aggregate millions of patient records, genomic sequences, and behavioral traces. Gene editing technologies allow precise modification of living organisms. Each of these capabilities creates new ethical questions that older RCR curricula did not anticipate. An instructor who teaches only the classic cases of data fabrication and plagiarism will leave students unprepared for the integrity decisions they will face in modern laboratories.
The practical problem for an RCR instructor is curriculum design. You need to know which emerging topics matter most, how to teach them within existing time constraints, and where to find authoritative guidance. This article offers a module outline, teaching resources, and assessment strategies grounded in the published standards of major research organizations.
The Core Domains of Responsible Research
Before adding new content, an instructor should understand the existing structure of RCR training. The NIH Grants and Funding pages describe the expectations for researchers who receive federal support, including training in responsible research practices. The COPE Core Practices organize publication ethics into several domains that map well onto RCR instruction.
Authorship and Publication Ethics
Authorship is the most visible domain of research integrity. The COPE Core Practices state that authorship should be based on substantial intellectual contribution, and that all authors must approve the final version of a work. AI tools complicate this standard. If an AI system drafts a paragraph, is it an author? The consensus among research organizations is that AI cannot be an author because it cannot take responsibility for the work. But the question of how to disclose AI assistance remains open in many fields.
An RCR instructor should teach students to ask three questions about any contribution. First, did the person make a substantial intellectual contribution to the design, analysis, or interpretation? Second, can the person take public responsibility for the content? Third, does the person approve the final version? These questions come from the COPE authorship guidance and apply equally to human and machine contributions.
Data Management and Sharing
Data management has become a formal requirement for NIH-funded research. The NIH Data Management and Sharing Policy requires researchers to submit a data management plan with their applications and to share scientific data at the time of publication. This policy creates concrete obligations that RCR students must understand.
AI research creates special data challenges. Training datasets may contain sensitive information about individuals. Models may memorize and reproduce personal data from their training sets. Researchers must plan for these risks when they write their data management plans. The NIH policy provides a framework for thinking about data sharing, but it does not answer every question about AI-specific data risks. Instructors should teach students to ask what data their model will see, what it might memorize, and what obligations they have to protect the people or animals represented in that data.
Research Misconduct and Questionable Practices
Research misconduct has a specific definition in US federal policy. It covers fabrication, falsification, and plagiarism. The COPE Core Practices describe how journals should handle allegations of misconduct. AI introduces new forms of misconduct that do not fit neatly into the traditional categories.
For example, a researcher might use an AI tool to generate a Western blot image that looks like real data. This is falsification, but the researcher might claim they did not know the tool would produce a misleading image. An RCR instructor should teach students that ignorance of a tool's behavior is not a defense. Researchers are responsible for understanding the tools they use and for verifying that their outputs are accurate.
The AI Ethics Landscape for Research
AI ethics is a broad field, but RCR instructors do not need to teach the entire philosophy of AI. They need to teach the specific ethical obligations that AI creates for researchers. These obligations fall into several categories.
Data Privacy and Protection
AI systems often require large datasets for training. These datasets may contain personal information, health records, or other sensitive data. The NIH Data Management and Sharing Policy requires researchers to consider privacy and confidentiality when they plan their data sharing. The same considerations apply to AI training data.
An RCR instructor should teach students to ask three questions about any dataset they plan to use. What information does the dataset contain? Who is represented in the dataset? What are the risks of re-identification? These questions are not unique to AI, but AI makes them more urgent because models can extract patterns that humans cannot see.
Algorithmic Bias and Fairness
AI systems can produce biased results when their training data is not representative. A model trained on one population may not work for another. This is a research integrity issue because it affects the validity of the research findings. The Research Methods Resources gateway provides access to authoritative biomedical references that discuss study design and validity. An RCR instructor can use these resources to teach students how to evaluate whether an AI model is valid for the population it will be used on.
Accountability for AI Outputs
When an AI system makes an error, who is responsible? The researcher who used the system, the developer who built it, or the institution that deployed it? The COPE Core Practices emphasize that researchers must be accountable for their work. This accountability extends to the tools they use. A researcher cannot blame an AI for a mistake that they should have caught.
Big Data Ethics in Research
Big data research involves the collection and analysis of datasets that are too large for traditional methods. These datasets often come from electronic health records, social media, or sensor networks. The ethical issues in big data research are similar to those in AI research, but they have their own specific features.
Data Provenance and Quality
Big data is often repurposed from other contexts. A dataset collected for clinical care might be reused for research. A dataset collected for marketing might be reused for public health. This reuse creates questions about data provenance and quality. The EQUATOR Network provides reporting guidelines that help researchers describe their data and methods transparently. An RCR instructor should teach students to document where their data came from and what limitations it has.
Informed Consent in the Age of Big Data
Traditional informed consent asks participants to agree to a specific study. Big data research often uses data that was collected for other purposes. The question of whether this reuse requires new consent is a matter of ongoing debate. The NIH Grants and Funding pages describe the human subjects protections that apply to federally funded research. An RCR instructor should teach students to consult their institutional review board when they plan to reuse data.
Data Security and Breach Response
Big data creates larger targets for security breaches. A single breach can expose the data of thousands or millions of people. The NIH Data Management and Sharing Policy requires researchers to describe their security measures in their data management plans. An RCR instructor should teach students to think about security as a research integrity issue, beyond an IT issue.
Gene Editing and Research Ethics
Gene editing technologies such as CRISPR have created new possibilities for research and new ethical questions. The [RCR instructor] must address these questions because they affect the design, conduct, and reporting of research.
Responsible Design of Gene Editing Studies
Gene editing research requires careful design to avoid unintended effects. The Research Methods Resources gateway provides access to biomedical research methods that can help researchers design rigorous studies. An RCR instructor should teach students to consider the potential off-target effects of gene editing and to plan for their detection.
Reporting and Reproducibility
Gene editing studies must be reported transparently so that other researchers can reproduce them. The EQUATOR Network provides reporting guidelines for many study types. An RCR instructor should teach students to select the appropriate reporting guideline for their study design and to follow it.
Ethical Oversight of Gene Editing
Gene editing research is subject to institutional oversight. The NIH Grants and Funding pages describe the review processes that apply to federally funded research. An RCR instructor should teach students to understand the oversight requirements that apply to their work and to seek guidance when they are uncertain.
Building an RCR Module on Emerging Technologies
An RCR instructor who wants to update their curriculum can use the following module outline. The module is designed to fit into a single session or to be expanded into a series of sessions.
Module Objectives
The module should have clear learning objectives. By the end of the module, students should be able to identify the ethical issues that AI, big data, and gene editing create for research. They should be able to apply existing RCR principles to new technologies. They should be able to find authoritative guidance when they face an unfamiliar ethical question.
Session One: AI and Research Integrity
The first session should focus on AI. The instructor should introduce the COPE Core Practices and explain how they apply to AI-assisted research. The instructor should use case studies to illustrate the issues. For example, a case study might describe a researcher who uses an AI to draft a manuscript and does not disclose the AI use. The instructor should ask students to discuss whether this is acceptable and what the researcher should have done.
Session Two: Big Data and Research Ethics
The second session should focus on big data. The instructor should introduce the NIH Data Management and Sharing Policy and explain its requirements. The instructor should discuss the ethical issues of data reuse, privacy, and security. The instructor should use case studies to illustrate these issues.
Session Three: Gene Editing and Research Ethics
The third session should focus on gene editing. The instructor should introduce the ethical questions that gene editing raises. The instructor should discuss the design, reporting, and oversight of gene editing research. The instructor should use case studies to illustrate these issues.
Session Four: Integration and Application
The fourth session should integrate the material from the previous sessions. The instructor should present a complex case that involves AI, big data, and gene editing. The instructor should ask students to apply the principles they have learned to the case. The instructor should also ask students to identify the resources they would use to resolve the ethical questions.
Teaching Resources for RCR Instructors
An RCR instructor needs reliable resources to support their teaching. The following resources are authoritative and freely available.
The NIH Grants and Funding Pages
The NIH Grants and Funding pages provide official information about NIH policies and procedures. An instructor can use these pages to teach students about the requirements that apply to federally funded research. The pages are updated regularly, so the instructor should check them for the latest information.
The NIH Data Management and Sharing Policy
The NIH Data Management and Sharing Policy is a specific policy that applies to NIH-funded research. The instructor should teach students about this policy because it affects how they plan and conduct their research. The policy also provides a framework for thinking about data ethics.
The COPE Core Practices
The COPE Core Practices describe the ethical expectations for publication. The instructor should use these practices to teach students about authorship, peer review, and publication ethics. The practices are relevant to all research fields.
The EQUATOR Network
The EQUATOR Network provides reporting guidelines for research studies. The instructor should teach students to use these guidelines to report their research transparently. The guidelines are especially important for AI and gene editing research, where the methods are complex.
The Research Methods Resources Gateway
The Research Methods Resources gateway provides access to authoritative biomedical books and references. The instructor can use these resources to teach students about research design and validity. The resources are particularly useful for teaching about AI and gene editing research.
ORCID for Researchers
The ORCID for Researchers pages describe how researchers can maintain a persistent identifier for their work. The instructor should teach students to use ORCID to manage their research identity. This is important for authorship and for tracking research contributions.
At a Glance: Emerging Technology Ethics in RCR
The following table summarizes the key ethical issues that AI, big data, and gene editing raise for research. The table also identifies the relevant authoritative source for each issue.
| Technology | Key Ethical Issue | Relevant Source |
|---|---|---|
| AI | Authorship and disclosure of AI contributions | COPE Core Practices |
| AI | Data privacy and security in AI training | NIH Data Management and Sharing Policy |
| Big Data | Data provenance and reuse | EQUATOR Network |
| Big Data | Informed consent for data reuse | NIH Grants and Funding |
| Gene Editing | Design and reporting of gene editing studies | Research Methods Resources |
| Gene Editing | Institutional oversight and review | NIH Grants and Funding |
Practical Implementation Steps for RCR Instructors
An RCR instructor can follow these steps to update their curriculum.
Step 1: Review the Existing Curriculum
The instructor should review their current RCR curriculum to identify the topics that are already covered. The instructor should also identify the topics that are missing. The instructor should use the COPE Core Practices and the NIH policies as a checklist.
Step 2: Identify the Emerging Technologies That Matter
The instructor should identify the emerging technologies that are most relevant to their students. The instructor should consider the research fields of their students and the technologies that are commonly used in those fields. The instructor should also consider the technologies that are likely to become important in the near future.
Step 3: Develop the Module Content
The instructor should develop the module content using the outline provided in this article. The instructor should use the authoritative sources to support the content. The instructor should also develop case studies that are relevant to their students.
Step 4: Deliver the Module
The instructor should deliver the module to their students. The instructor should use a variety of teaching methods, including lectures, discussions, and case studies. The instructor should also assess the students' understanding of the material.
Step 5: Evaluate and Update the Module
The instructor should evaluate the module after delivery. The instructor should collect feedback from the students and use it to improve the module. The instructor should also update the module as new technologies and new policies emerge.
Records and Measurements for RCR Training
An RCR instructor should keep records of their training activities. The records should include the date of the training, the topics covered, and the names of the participants. The records should also include the evaluation results.
The instructor should measure the effectiveness of the training. The instructor can use pre- and post-training assessments to measure the students' knowledge. The instructor can also use surveys to measure the students' satisfaction with the training.
The instructor should use the records and measurements to improve the training. The instructor should identify the topics that students found difficult and the topics that students found useful. The instructor should use this information to adjust the training.
Common Failure Patterns in RCR Training
RCR training can fail in several ways. The instructor should be aware of these failure patterns and take steps to avoid them.
Failure Pattern 1: Treating Emerging Technologies as a Separate Topic
One common failure is to treat emerging technologies as a separate topic that is unrelated to the core RCR principles. This approach fails because it does not help students see how the principles apply to new technologies. The instructor should integrate the emerging technologies into the core RCR topics.
Failure Pattern 2: Focusing on the Technology Instead of the Ethics
Another common failure is to focus on the technology itself instead of the ethical issues. The instructor should focus on the ethical issues and use the technology as a context for discussing them.
Failure Pattern 3: Using Outdated Information
A third common failure is to use outdated information. The instructor should use the most current information from the authoritative sources. The instructor should also update the module as new policies and technologies emerge.
Failure Pattern 4: Not Providing Practical Guidance
A fourth common failure is to provide only theoretical guidance without practical guidance. The instructor should provide practical guidance that students can use in their own research. The instructor should also provide the students with the resources they need to find the guidance.
Limitations and Professional Escalation Criteria
The guidance in this article has limitations. The instructor should be aware of these limitations and should escalate to the appropriate authority when necessary.
Limitations of the Module
The module is not a substitute for the institutional policies and the institutional review board. The instructor should teach students to follow their own institutional policies and to consult their institutional review board when they have questions.
The module is not a substitute for the legal and regulatory requirements. The instructor should teach students to comply with the legal and regulatory requirements that apply to their research.
Professional Escalation Criteria
The instructor should escalate to the appropriate authority when they encounter a situation that is beyond their expertise. The instructor should escalate to the institutional review board when they have questions about human subjects research. The instructor should escalate to the institutional animal care and use committee when they have questions about animal research. The instructor should escalate to the research integrity officer when they have questions about research misconduct.
Welfare and Safety Context
The ethical issues raised by AI, big data, and gene editing have welfare and safety implications. The instructor should teach students to consider the welfare and safety of the people and animals who are affected by their research.
Human Welfare and Safety
The instructor should teach students to consider the welfare and safety of the people who are represented in their data. The instructor should teach students to protect the privacy and confidentiality of the people in their data. The instructor should also teach students to consider the potential harms of their research.
Animal Welfare and Safety
The instructor should teach students to consider the welfare and safety of the animals that are used in their research. The instructor should teach students to follow the institutional animal care and use committee requirements. The instructor should also teach students to consider the potential harms of their research to the animals.
Frequently Asked Questions
How can an RCR instructor integrate AI ethics into an existing curriculum?
An RCR instructor can integrate AI ethics by mapping AI issues to the existing RCR domains. For example, the instructor can discuss AI authorship in the authorship module and AI data privacy in the data management module. The instructor can also add a dedicated session on AI ethics, as described in the module outline in this article.
What is the role of the COPE Core Practices in RCR training?
The COPE Core Practices provide a framework for teaching publication ethics. The instructor can use the practices to teach students about authorship, peer review, data, conflicts, misconduct, and publication ethics. The practices are relevant to all research fields.
How does the NIH Data Management and Sharing Policy affect AI research?
The NIH Data Management and Sharing Policy requires researchers to plan for data management and sharing. This policy applies to AI research because AI research often uses large datasets. The instructor should teach students to consider the data management and sharing requirements when they plan their AI research.
What is the role of the EQUATOR Network in RCR training?
The EQUATOR Network provides reporting guidelines for research studies. The instructor should teach students to use these guidelines to report their research transparently. The guidelines are particularly important for AI and gene editing research, where the methods are complex.
How can I teach students about the ethical use of gene editing in research?
The instructor can teach students about the ethical use of gene editing by discussing the design, reporting, and oversight of gene editing research. The instructor can use the NIH Grants and Funding pages to teach students about the review processes that apply to gene editing research.
What should I do if I have a question about the ethics of a specific research project?
The instructor should consult the authoritative sources and the institutional policies. The instructor should also consult the institutional review board or the research integrity officer. The instructor should not rely on the module alone to answer specific questions.
How can I keep my RCR training current?
The instructor should regularly review the authoritative sources and update the module. The instructor should also monitor the emerging technologies and the new policies. The instructor should also collect feedback from the students and use it to improve the training.
What is the role of ORCID in research ethics?
The ORCID for Researchers pages describe how researchers can maintain a persistent identifier for their work. The instructor should teach students to use ORCID to manage their research identity. This is important for tracking the research and for ensuring that the research is properly attributed.
Using the Evidence
| Source | Best use in this topic | Important limitation |
|---|---|---|
| Research Methods Resources | official guidance | Check the linked page for current local requirements |
| EQUATOR Network | official guidance | Check the linked page for current local requirements |
| Core Practices | official guidance | Check the linked page for current local requirements |
Related Articles
- Ethical Research Practices: A Guide to Responsible Conduct
- Ethical Data Collection in Research: Principles and Best Practices
- Ethical Issues in Data Collection: Navigating Privacy and Consent
- Ethical Considerations in Computational Genomics: Data Privacy and Responsible Use
- The Tuskegee Syphilis Study: Ethical Violations and Lessons for Modern Research
References and Further Reading
- Research Methods Resources. National Library of Medicine.
- EQUATOR Network. EQUATOR Network.
- Core Practices. Committee on Publication Ethics.
- NIH Grants and Funding. National Institutes of Health.
- ORCID for Researchers. ORCID.
- Data Management and Sharing Policy. National Institutes of Health.
- NCBI Literature Resources. National Center for Biotechnology Information.
- PubMed. National Library of Medicine.
This article is educational and does not replace institutional policy, professional advice, or applicable safety and regulatory requirements.