‘Junk science’ fabricated by AI floods Google Scholar, researchers warn

AI-generated research is a threat both to society’s knowledge and to public trust in science. This is the conclusion reached by researchers at the Swedish School of Library and Information Sciences at Borås University in Sweden, who recently identified more than 100 papers suspected of being generated by AI in the Google Scholar search engine.

Their research shows that search engine Google Scholar is being flooded with fake scientific papers generated by AI. Research findings on AI-fabricated junk science mean that fake science is now available and can be widely disseminated at a much lower cost to malicious actors. The study was published in the Harvard Kennedy School Misinformation Review.

According to Jutta Haider and Björn Ekström from the Swedish School of Library and Information Sciences, who are leading the research along with Christopher Söderström from Lund University and Malte Rödl from the Swedish University of Agriculture, this They say it poses a danger to both communities. Science.

Increased risk of evidence hacking

One of the main concerns about AI-generated research is the increased risk of evidence hacking, meaning that fake research can be used for strategic manipulation.

“As research generated by AI becomes more widespread on search engines, the risk of so-called ‘evidence hacking’ increases significantly. It has the potential to have tangible consequences,” Ekström said. in Library and Information Science.

Researchers involved in this study have already seen these problematic papers spread to other parts of the research infrastructure, including the web, various archives, and social media. It spreads quickly, and problematic papers are visible on Google Scholar. Even if the article is taken down, it has already had time to spread and is at risk of continuing to spread.

Additionally, AI-generated research poses challenges to an already challenged peer review system.

Higher demands on information literacy

The fact that AI-generated surveys are pervasive in search engine databases places higher demands on people who participate in surveys online.

“If you can’t trust that the studies you read are real, you run the risk of making decisions based on misinformation. But this is as much a media and information literacy issue as it is a scientific misconduct issue.” said Professor Haider. Library and Information Science.

She points out that Google Scholar is not an academic database. The search engine is easy to use and fast, but it lacks quality assurance procedures. This is already a problem with regular Google search results, but it’s even more of a problem when it comes to making science accessible.

“People’s ability to decide which journals and publishers publish high-quality, reviewed research is critical to finding and determining what is trustworthy research, and is important for decision-making and opinion-forming. “This is of great importance to the world,” Haider concluded.