Jason Semprini was excited about his research on policies mandating that elementary school students receive the human papillomavirus vaccine. HPV causes most cases of cervical cancer, but counterintuitively, Semprini found that mandates don’t do that much to reduce the overall rate of cervical cancer in a population. That’s odd, but it's also not completely surprising—we know that mandating a vaccine can motivate some people to find ways to avoid it.

Semprini wrote up the study and submitted the manuscript to a journal, where it was sent out for peer review, a long-standing process through which other researchers in the field assess the validity of research and make a recommendation on whether or not it should be published. This is usually done anonymously and on a volunteer basis.

In this case, the reviewer did not share Semprini’s excitement. That may have stemmed from a misunderstanding of the study’s central message: The reviewer mistakenly thought Semprini was questioning whether the HPV vaccine itself prevents cervical cancer rather than studying the effectiveness of a policy meant to increase vaccination rates. “There’s a very big difference there,” said Semprini, who is a health economist at Des Moines University.

Read full article

Comments

Jason Semprini was excited about his research on policies mandating that elementary school students receive the human papillomavirus vaccine. HPV causes most cases of cervical cancer, but counterintuitively, Semprini found that mandates don’t do that much to reduce the overall rate of cervical cancer in a population. That’s odd, but it's also not completely surprising—we know that mandating a vaccine can motivate some people to find ways to avoid it.

Semprini wrote up the study and submitted the manuscript to a journal, where it was sent out for peer review, a long-standing process through which other researchers in the field assess the validity of research and make a recommendation on whether or not it should be published. This is usually done anonymously and on a volunteer basis.

In this case, the reviewer did not share Semprini’s excitement. That may have stemmed from a misunderstanding of the study’s central message: The reviewer mistakenly thought Semprini was questioning whether the HPV vaccine itself prevents cervical cancer rather than studying the effectiveness of a policy meant to increase vaccination rates. “There’s a very big difference there,” said Semprini, who is a health economist at Des Moines University.

Read full article

Comments

On Tuesday, the White House Office of Science and Technology Policy (OSTP) released a report entitled "Science: A New Golden Age," in which it lays out how it has viewed science, found it lacking, and believes the Trump administration is in the perfect position to fix things. It's a bit unexpected coming from an administration that has been proposing crippling funding cuts to research and trying to enable political appointees to terminate grants awarded based on scientific merit.

The report presents itself as the spiritual successor of "Science, the Endless Frontier," a policy document that laid out the case for government-funded science in the wake of World War II. The New Golden Age (SNGA) says that, while the concepts promoted by the original remain vital, the circumstances have changed such that we need major revisions to how the government is implementing things.

The result is an odd mix. It completely ignores or glosses over many things that the administration is doing to harm scientific progress. In some cases, it identifies issues that have already been discussed as problems within the scientific community, but misrepresents them. Elsewhere, it's a mixture of political grievances, ideas without a solid intellectual foundation, and an injection of Silicon Valley's perspective on innovation (Michael Kratsios, the director of the OSTP, formerly worked with Peter Thiel). As a result, it's unlikely to have anything like the impact of "Science, the Endless Frontier."

Read full article

Comments

In a pioneering stride towards maintaining the integrity of scientific literature, MDPI has fully implemented its proprietary artificial intelligence-powered system called Ethicality. This sophisticated tool is now embedded within the publishing workflow to automatically screen all submitted manuscripts, revolutionizing the approach to safeguarding the scientific record. Ethicality is engineered to operate continuously throughout the editorial process, offering a proactive framework that enhances the evaluation and decision-making capabilities of editors globally. Presently, it processes around 2,000 submissions daily, handling a vast and diverse array of academic contributions from across the world.

The development of Ethicality reflects the pressing need for automated solutions in the face of surging scholarly output and increasing demands for transparency, speed, and quality in academic publishing. Traditional manual screening methods, once the norm, are no longer adequate to detect sophisticated manipulations or emerging threats posed by advanced generative AI techniques. Its design integrates an end-to-end integrity layer that scrutinizes manuscripts from the moment they enter the editorial system, ensuring continuous vigilance rather than episodic checks. This paradigm shift marks a new era in editorial practices, combining the precision of AI with human expertise to set robust guardrails around the publication pipeline.

Dr. Milos Cuculovic, MDPI’s head of technology innovation, emphasizes that the role of AI within this context is not to replace editors but to augment their capabilities. Ethicality functions as a vigilant assistant, flagging potential ethical issues and inconsistencies while leaving final judgment to experienced professionals. This human-in-the-loop approach enables a balance where automation streamlines routine tasks without compromising critical oversight. Such a model helps preserve transparency and trust at scale, combating increasingly sophisticated tactics that jeopardize scientific authenticity, such as synthetic content generation, fabricated submissions, and manipulated peer review reports.

Ethicality’s analytical engine examines each submission comprehensively by dissecting fundamental manuscript components, including the title, abstract, author metadata, main body, and reference lists. Beyond textual elements, it also reviews peer review reports sent with the manuscripts, employing natural language processing techniques and machine learning algorithms tailored for scholarly content. This holistic evaluation identifies a wide spectrum of potential integrity breaches, ranging from paper mill-generated manuscripts and spurious authorship claims to subtle citation manipulations and irregular referencing patterns that may indicate gaming of metrics or unethical publication practices.

One of the system’s remarkable features lies in its ability to detect AI-generated or altered text, a rapidly emerging threat in the digital publication ecosystem. Leveraging large language models fine-tuned to recognize nuances and inconsistencies characteristic of synthetic content, Ethicality can pinpoint sections likely created or tampered with by artificial intelligence. This capability is crucial amid widespread concerns that generative AI tools could be exploited to fabricate scientific results or obfuscate fraudulent data, undermining trust in scholarly communication. The AI’s detection mechanisms, however, are continuously refined against real-world data to minimize false positives and maintain editorial efficiency.

In addition to content analysis, Ethicality collaborates with external software suites to cover domains beyond textual scrutiny. For instance, image manipulation detection is managed via integration with specialized tools like Proofig, which applies forensic algorithms to uncover alterations in scientific figures and illustrations. For plagiarism and text duplication, the system incorporates iThenticate services to cross-reference submissions with an extensive repository of published works. This multi-layered approach strengthens MDPI’s ability to uphold ethical standards by combining internal AI capabilities with trusted third-party solutions to comprehensively monitor manuscript authenticity.

The advent of such an automated system answers growing concerns about the sustainability of peer review in an era characterized by exponential growth in submission volumes. By automating repetitive and technical screening tasks, Ethicality frees up valuable editorial resources, allowing human experts to concentrate on evaluating the scientific merit and originality of manuscripts rather than being overwhelmed by manual integrity checks. This redistribution of effort ensures that editorial workflows can keep pace with demand without compromising quality or publication timelines, addressing one of the major bottlenecks in modern scholarly publishing.

Dr. Enric Sayas, the product owner of Ethicality, highlights that the publishing sector is entering a technological race where adversaries leverage advanced AI to generate deceptive manuscripts and fake review reports. Traditional detection tools alone are insufficient to counteract such sophistication. Only by deploying equally advanced AI systems, capable of multi-modal analysis encompassing textual, image, and metadata anomalies, can publishers effectively defend the credibility of scholarly records. Ethicality embodies this next-generation defense mechanism, combining deep learning models with domain-specific heuristics to expose fraudulent submissions before they can impact the scientific corpus.

Ethicality’s role extends beyond mere detection; it facilitates informed decision-making by generating risk signals that prioritize editorial scrutiny. This nuanced output classifies flagged cases according to severity and type of ethical concern, enabling editors to apply their judgment where most critical while bypassing lower-risk items. Such targeted intervention reduces unnecessary delays in the publication process and improves the overall efficiency of manuscript handling. The system’s iterative design also permits continuous learning from editor feedback, enhancing accuracy and adaptability as new manipulation techniques emerge.

The timing of Ethicality’s deployment coincides with an inflection point in scholarly publishing, where generative AI tools have made it increasingly feasible to produce lifelike but fraudulent scientific content. This proliferation necessitates automated defenses that not only identify outright fraud but also detect subtler anomalies such as inconsistent author data, suspicious referencing behavior, and peer review irregularities. By integrating these multifaceted analyses, MDPI demonstrates a comprehensive and forward-looking strategy aimed at preserving the integrity and credibility of academic knowledge dissemination.

Crucially, Ethicality exemplifies how technology can be harnessed responsibly within complex human ecosystems. It respects the indispensable role of expert editors by operating as a sentinel that brings potential issues to attention while leaving interpretive authority intact. This model upholds core values of transparency, accountability, and collaborative oversight, providing a blueprint for other publishers grappling with integrity challenges in the AI era. Moreover, it underscores the broader imperative for publishers to innovate continuously in response to evolving threats and technological opportunities, ensuring that scientific publishing remains a trusted pillar of global knowledge advancement.

As MDPI’s journey with Ethicality moves forward, continuous enhancements and iterations are planned, driven by real-world data and editorial collaboration. This adaptive approach ensures the system evolves alongside the changing landscape of academic publishing and the advent of new manipulative tactics. Ultimately, Ethicality stands as a testament to the potential of AI-powered tools to reinforce scientific integrity, protect the scholarly record, and empower editorial professionals in an increasingly complex and dynamic environment.

Subject of Research: Artificial intelligence applications in scholarly publishing and research integrity
News Publication Date: 2024
Image Credits: MDPI

Keywords

Academic publishing, research integrity, artificial intelligence, editorial workflow automation, AI-generated scientific content detection, paper mill detection, peer review screening, scholarly publishing technology, automated manuscript analysis, scientific record safeguarding