Using AI in research: An interview with Elsevier Senior Vice President Gemma Hersh — The most important thing is the reliability of underlying data

Elsevier is a major global academic publisher that in recent years has been providing databases such as Scopus, as well as AI services based on vast amounts of research paper data. The Science News spoke with Gemma Hersh, Senior Vice President of Global Academic and Government Sales, about the use of AI in academic fields.

Work quality is improving

— How will AI change research?

We believe that AI allows researchers to advance and accelerate their research. It has the potential to transform the research process, and we are seeing many examples of this around the world.

Our investigations show that the majority of researchers in Japan are already using AI tools. More than 80% of researchers said that generative AI would improve the time efficiency required for research. Additionally, about half said that generative AI would also improve the quality of researchers' work.

Right now, the way research is conducted is truly changing, and I believe we are at the beginning of a very exciting journey.

Japan is positive

— Is Japan more advanced in AI adoption compared with other countries?

Our research to date shows that the Asia-Pacific region—namely China, Japan, and other countries—is more open to AI compared with other countries, such as the United States. This means Japan has the opportunity to be an early adopter, embracing AI and being among the first to harness its potential.

When we speak with institutions in other countries, some are cautious, some are taking a wait-and-see approach, with others wanting to assess the situation first. On the other hand, the Japanese government and funding agencies are willing to provide financial support for AI.

In one of our recent reports, many researchers indicated that they feel AI training is needed and that AI which can be integrated into their research infrastructure and workflows is necessary.

Among various countries, Japan is particularly positive. MEXT (the Ministry of Education, Culture, Sports, Science and Technology) is advancing AI literacy programs, and integrating AI into workflows is one of the priority areas in its AI strategy.

However, there are important points to consider in order to accelerate this movement. Trust is important for research tools. The foundation—that the quality of the underlying content and data is reliable—is extremely important. One of the key principles for this is transparency: knowing where the information and data come from, and what sources were referenced.

A tool to support researchers

— Regarding concerns that researchers might become dependent on AI

The important point is that AI is not meant to replace researchers. AI is meant to support critical thinking.

It is used to speed up researchers' own work, make tasks easier, or help them gain knowledge and insights more quickly. Researchers must always examine the output from the perspective of critical thinking.

That is why it is important to focus on AI tools for researchers. AI tools for researchers are designed to speed up work output, helping save time while also providing functions that support critical thinking or offer guidance.

As AI becomes more widely used in research, the principle of transparency becomes even more important. For example, when using an agentic AI to conduct research, I think it is very important to lay out and clarify from the beginning, at every step, how researchers interpret and use the output from the AI, what they cite, and where the sources come from.