Home > Research Integrity > Responsible Use of Generative AI in Research at HKU

Responsible Use of Generative AI in Research at HKU

Recent developments in generative artificial intelligence (GenAI) technologies are revolutionising the research landscape by enabling innovative approaches to data analysis, content creation, hypothesis generation, and problem-solving. Concerns regarding data quality, bias, ownership, and potential misuse necessitate responsible and ethical use to uphold research integrity, safeguard participants’ rights, and maintain societal trust.
 
HKU is dedicated to promoting the highest standards of research integrity and is committed to fostering responsible GenAI practices and promoting best practices in the adoption of GenAI among researchers, students, supervisors and other staff in the research community. In response, the University Research Committee (URC) set up a Task Force to formulate an institutional policy on GenAI in research. Following a comprehensive study and stakeholder consultation between May and September 2025 through a mixer event, surveys, and a circulation of draft guidelines to the Graduate School and all Faculties, the University issued the Guidelines on the Responsible Use of Generative AI in Research at The University of Hong Kong which was approved by the Senate at its meeting on September 23, 2025.
 
 

What is GenAI?

GenAI refers to machine learning models capable of generating human-like text, images, audio, or other media based on input prompts. These tools can assist with literature reviews, data interpretation, report drafting, visualisations, and other applications. Examples include:
  • ChatGPT for natural language generation
  • DALL·E for image creation
  • Audio and video synthesis tools
 

General Principles of Responsible Use

1. Research Integrity
  • Avoid misleading information and ensure AI-generated content is accurate and do not violate research integrity principles, including fabrication, falsification, or misconduct.
  • Be aware of potential biases in GenAI outputs and interpret results critically.
2. Data Validity and Quality Control
  • Critically verify and interpret GenAI-generated content.
  • Use human judgment and domain expertise to assess GenAI outputs.
  • Do not rely solely on GenAI-generated content for research conclusions or publications.
3. Transparency and Disclosure
  • Clearly disclose GenAI tool usage in research outputs, publications, or presentations, for example the type of GenAI used, data sources, and potential limitations.
  • Maintain detailed documentation of GenAI prompts, outputs, and decision-making processes involving GenAI.
4. Data Privacy and Confidentiality
  • Ensure input data used with GenAI complies with data protection laws and University policies.
  • Avoid sharing sensitive, confidential, or proprietary information with GenAI platforms.
5. Responsible Data Management
  • Store and manage all data used and generated by GenAI securely.
  • Maintain comprehensive records of GenAI inputs and outputs for reproducibility, compliance, and accountability.
6. Respect Intellectual Property and Ownership
  • Ensure that the use of GenAI does not infringe on existing rights or compromise research outputs ownership.
7. Accountability
  • Researchers are responsible for the outputs generated by GenAI and their implications.
 

Best Practices for Responsible Use

  • Engage in Ongoing Education and Training: Regularly and actively update knowledge on GenAI capabilities, ethical considerations, and limitations, as well as stay updated on the latest developments in GenAI.
  • Provide Training and Supervision: Identify and design GenAI-supported and GenAI-free tasks for research students and trainees. 
  • Document Usage: Keep detailed records of how and when GenAI tools are used in the research process. 
  • Disclose Usage: Clearly acknowledge GenAI contributions explicitly in research records, reports, and publications. 
  • Limit Over-Reliance: Use GenAI as a support tool, not a substitute for critical analysis and human expertise. 
 

Other Related HKU Policies

 

References and Further Reading

 
----------------------------------------------------------------------------------------------------------------------------
 
Notes
This webpage aims to foster a culture of responsible and ethical use of GenAI in research, in alignment with HKU's overarching policies on research integrity and data management. It does not replace or supersede the standards set out in the Guidelines on the Responsible Use of Generative AI in Research at The University of Hong Kong. This webpage should also be read in conjunction with other institutional policies and regulations.