The new European code of scientific ethics also affects artificial intelligence

The Association of European Scientific Academies (All European Academies, ALLEA) updated its publication of the European Code of Conduct for Research Integrity, which sets out ethical rules for scientific research in 2017, to meet the challenges of recent years. The new version entailed a number of circumstances that, at the time of preparation of the previous version, did not affect the practice of science as strongly as it is today. Chief among them is the emergence of artificial intelligence in the daily lives of researchers, but as assessed by ALLEA, aspects of data protection and elimination of discrimination, ensuring diversity and fairness of science, proliferation of open access journals and management of related breaches are equally important. Laszló Fesusa regular member of the Hungarian Academy of Sciences, Emeritus Professor at the Institute of Biochemistry and Molecular Biology at the University of Debrecen, and member of the Permanent Working Group on Scientific Ethics at ALLEA, asked about mta.huwhich currently threatens the integrity of research.

Why was it necessary to create the European Code of Scientific Ethics?

European Code of Conduct European Research Integrity The Research Code was developed by the European Research Community to regulate its own work. For more than two decades now, it has been widely seen as a guideline, while national codes of conduct also exist. In 2010, the Hungarian Academy of Sciences created the Hungarian Code of Scientific Ethics, which is still valid today, and which will soon also need to be amended, to adapt to new conditions and new European law. Since the ethical regulation of research is very different in the countries of the continent, the demand for a blog at the European level was naturally formulated. This has also become of practical importance since, since 2017, the European Code of Scientific Ethics should be considered the primary reference when awarding EU grant money and when concluding support contracts. In this sense, this document provides a strong research ethics framework for all European institutions and researchers. The 2017 Hungarian version has also been published, and a Hungarian translation of the updated manuscript will also be prepared.

What made it necessary to update the previous 2017 version of the European Code of Ethics?

In recent years, great changes have taken place in the practice and technology of scientific research. Data protection, access and data management aspects are becoming more prominent. In addition, the use of social media in science communication has exploded in the past five years. Questions and inconsistencies have also arisen regarding publication in open access journals, forcing us to reformulate certain elements of the code of ethics with a different focus. ALLEA also highlights that in light of these developments, the responsibility of institutions has become even more important in ensuring a research culture based on the preservation of scientific research integrity, respect for generally accepted codes of conduct and good research practices. Consciously maintained and developed.

What principles should good research practices be based on?

The principles of scientific ethics formulated in the Code have not changed much compared to the previous edition, at most their wording has been revised. The four basic principles that should influence the work of all researchers are reliability, honesty, respect (to the research subject, its participants, the community, the ecosystem, and colleagues), and accountability. These four aspects perfectly summarize the ethical expectations that guide researchers. Good research and code research is detailed based on basic principles. Forms of violations of research obligations and unacceptable research activities are then described in the document, and finally the general rules of procedure to be followed in the event of a breach of research integrity are described.

Artificial intelligence is dominating science and technology news these days. What is the impact of this field on scientific integrity?

Artificial Intelligence (AI) was completely absent from the 2017 Code of Scientific Ethics, as a new situation in this regard has recently emerged. The code asserts that there is a place for AI in research practice, and researchers can use many powerful tools and third-party services capable of automation, even when preparing a publication. However, there is one very important condition: the use of artificial intelligence must be clearly indicated when results, data and texts are published. In other words, concealing the use of AI in communicating scientific findings to one is included among behaviors that violate researcher integrity and good practice.

The integrity of the publication is threatened not only by the unethical use of artificial intelligence, but also by so-called predatory magazines, which usually publish almost anything for money without any kind of standard oversight. How do we deal with the shortcomings of open access publishing?

There are approximately 40 thousand scientific journals in the world, about half of which are open. Soon after the start of the open access movement, so-called predatory open access journals emerged, which only seek to maximize publishing fees and profits, and thus publish almost anything quickly and without quality control. It was expected that there would be people who would take advantage of the opportunities inherent in open access to earn easy money and make extra profit. However, it was not expected that predatory activity would become so large-scale, and the resulting situation would become so dangerous. According to the databases, from 15,000 to 16,000 scientific journals can be classified as predators, acting with increasingly sophisticated methods, while a large number of researchers are not aware of the danger lurking in them. The research community must fight in every possible way to eradicate predatory phenomena. The European Code of Scientific Ethics can only classify the creation and support of journals and publishers that threaten research quality control, as well as deliberate publication in predatory journals, as unacceptable research practices. In addition, the new version of the code already mentions the Predator conferences that have appeared in recent years. These conferences are also based on fraud: they are advertised with the participation of famous researchers, but when researchers who pay participation fees go there, they are faced with the fact that the real event does not even remotely resemble the one promised in the advertisements.

To what extent can researchers be considered victims and to what extent are they complicit in the activities of predatory journals?

This section of the current code text has changed precisely in that it specifically emphasizes the responsibility of researchers: it marks as unethical the intentional use of these journals in order to aid scientific advancement, win bid money, and obtain scholarly qualifications. Such activity not only paints a false picture of individual results, but also undermines the quality assurance of science as a whole. It is difficult because it is not always possible to clearly identify which journals are classified as predatory and which are not, and there may be disagreements among researchers on this subject. The newly formed working committee of the Transportation Authority will soon prepare a report and submit a proposal in this regard. Accusations of predatory behavior are raised primarily in open access magazines, because subscription magazines will lose subscribers if the quality control they exercise is questioned. It can be a good guide for evaluating an open access journal if we check based on its title whether it is in one of the databases that list such journals based on quality criteria, for example in the directory List of Open Access Journals Verified by the International Research Society .

Does science face more serious ethical problems today than it did in the past? Is the integrity of researchers in a worse state because of new technological possibilities, or is the ongoing abuse or fraud becoming more apparent?

Traditionally, we have considered plagiarism, forgery, and forgery (when an author fabricates non-existent research results) among the most serious violations of scientific ethics. As a result of technological development, their frequency has changed, and some have become rarer, while others are more frequent (more noticeable). Plagiarism, for example, is on the decline, because text-checking methods (also aided by artificial intelligence) are now very effective at recognizing whether a text was created by plagiarizing another researcher’s work. Plagiarism checking is now a routine component of the evaluation of dissertations and manuscripts submitted for publication. However, forgeries and forgeries still occur regularly. If this is revealed, it usually results in a major scandal, and institutional, and sometimes criminal, proceedings are initiated against the researchers. I don’t think there are many such breaches, but because of the high-profile news stories surrounding the breaching researchers, we might get the impression that they are pretty common.

This is the most egregious scientific fraud. But how lenient are researchers with normal cases?

There are indeed research practices that may not be considered seriously unethical, but which nonetheless distort published scientific results and information. According to surveys, there are many of them, and they are more common, but at the same time less visible parts of the mass of violations of scientific ethics. These include, for example, the selective selection of results favorable to the researcher by excluding the less convincing parts, careless handling of statistical analysis, and manipulation of image data. 15-20 percent of researchers who answered anonymous surveys admitted that they sometimes distorted results in order to successfully publish them. It also often happens that the results of competing researchers are deliberately cited in the article, while it is full of their own references. Emeritus authorship is still common, when someone unrelated to the particular research still expects to be listed among the authors. In the manuscript review process, it also happens that the reviewer does not indicate that he or she belongs to a competing research group, and then deliberately hinders acceptance of the manuscript while trying to publish the results. It is an important task for the research community to do everything possible to eliminate and prevent unacceptable research practices listed in the updated code through training in scientific ethics, setting personal example in research groups, and mentoring, given the credibility of scientific research. Activity in the eyes of society largely depends on this.