5 ways to use AI ethically in research publications


The discovery of artificial intelligence ( AI) aims to create technology that can imitate human cognitive abilities. This includes learning new things, translating languages, processing information, solving and analyzing problems, and identifying patterns and trends in a phenomenon .

In the world of research, AI can help researchers find relevant information more quickly and easily, analyze data, translate manuscripts, summarize scientific articles, and develop models and simulations to understand complex phenomena. In fact, AI can help researchers perform experiments that were previously impossible for cost, time, or security reasons.

However, concerns regarding the misuse of AI in research publications have formed among some research groups and academics. They believe that AI is vulnerable to data falsification, manipulation of research results, plagiarism, and even theft of other people’s research. This will of course have a serious impact on trust in the authenticity of ideas and the quality of publications.

Risks of using AI in research publications

Several examples of AI-based conversational engines, such as Gemini from Google and ChatGPT from OpenAI, have the ability to find references and relevant data quickly and efficiently on the Internet.

In fact, users can also order AI to automatically conclude search result text without changing the original meaning . This ability can help researchers and writers in the stages of literature review to data interpretation.

However, this convenience risks “blunting” the researchers’ ability to read and think critically in the brainstorming process to finalize research. In fact, this ability is what differentiates the role of researchers which cannot be replaced by AI.

Ethics of using AI in research

The risks above demonstrate why it is important for researchers to understand and apply the ethical principles of using AI in research publications. So that AI can be used responsibly, fairly, and does not replace human research abilities.

1. Express ideas and research results in a distinctive writing style.

Even though you use AI, make sure you process and rewrite the search result information in your own writing style. AI technology can indeed help you improve grammar, but AI cannot replace writing creativity, critical thinking, and analytical processes.

A distinctive writing style shows our unique thoughts and ideas that AI cannot imitate. In addition, having a unique writing style can help build trust and connection between us as researchers and readers.

2. Mention that you use AI in research publications.

Even though you have rewritten it in your own writing style, still reveal the role of AI in the preparation of your publication manuscript. Clearly state whether AI played a role in preparing the background, methods, or discussion of the manuscript. This way, readers know that you are committed to using AI responsibly and transparently .

This practice is common. In fact, you can find several journals that require mention of the use of AI in research methods if the research involves AI in its preparation.

3. Check again where the AI ​​gets the information from.

When commanded, the AI ​​search engine can attach references such as website links, books, or scientific articles. However, sometimes AI search results return incorrect links or incorrect book names due to non-factual and limited data sources.

If you copy this information without checking it first, you risk committing plagiarism . Therefore, check again the reference recommendations provided by AI via search engines on the internet. This can help you find more valid sources.

By including appropriate references, you not only show your commitment to the values ​​of integrity and respect for other people’s work, but also convince readers that the writing you present is accurate and credible even though it uses AI assistance.

4. Select the appropriate AI.

Oliver Grünvogel , an expert in AI applications for Life Science , emphasizes the importance of choosing the right AI for a specific task, such as correcting grammar, creating summaries, or translating text. This is because AI systems with specific tasks are trained using data collection patterns according to their function, thereby producing more accurate outcomes or predictions and minimizing bias in research results.

Oliver also mentioned several examples of AI-based programs that researchers can use for writing publications, such as Paperpal, Text Blaze, Grammarly, and Writier. Meanwhile, for literature review tasks, Oliver recommends several AI applications such as R Discovery, Scite AI, Connected Papers, Research Rabbit, and Journal Finder.

5. Pay attention to how your data will be used and stored.

If you use AI to process research data, it is important to pay attention to privacy aspects. To increase privacy so that data is more difficult to hack or access, you can do things like, 1) Choose AI that does not store or process data in one central location, 2) Provide minimal or appropriate information necessary to complete the task, 3) Avoid AI that collects unnecessary personal data, such as personal information about research respondents or medical records, and 4) Avoid using AI that does not provide an explanation of how the data will be processed and a clear explanation of how it works.

Apart from that, setting privacy settings for each AI application we use can also help maintain data privacy .

Don’t lose to AI

Implementing ethical use of AI alone is not enough for a researcher. Researchers also have a responsibility regarding self-development so that they are not left behind amidst the rapid progress of AI technology. Researchers are required to sharpen their abilities in writing, collaborating, solving problems and analyzing data, thinking critically and creatively, and being able to understand programming languages.

In addition, it is time for educational institution ethics committees to start considering the ‘existence of AI use’ aspect in assessing the feasibility of research.

Author Bios: Shinta Noviar Unicha Pearls is a Research Assistant for Tuberculosis Studies, The Center for Tropical Medicine at Gadjah Mada University and Dito Aryo Prabowo is a Lecturer, Faculty of Psychology at Diponegoro University