The State of Scientific Publishing: Assessing Trends, Emerging Issues, and Policy Considerations

The Subcommittee on Investigations and Oversight will come to order. Without objection, the chair is authorized to declare recess of the subcommittee at any time. Welcome to today's hearing entitled, The State of Scientific Publishing: Assessing Trends, Emerging Issues, and Policy Considerations. I recognize myself for five minutes for an opening statement. Good morning and thank you for attending today. We are addressing a challenge that sits at the heart of the American scientific leadership, a challenge that demands honesty and scrutiny of the systems we rely on to produce and validate knowledge. Science advances through trust. Everything we do rests on the assumption that underlying work is real, rigorous, and honestly reported. I know the importance of this firsthand. From my time in medical school to practicing in the emergency room, my career has been based on published research and its credibility and its peer review process. Our actions are based on trust. Be it when a policymaker acts on published evidence or when a federal agency directs billions, if not trillions, of dollars in funding to specific research. That foundation is under threat. And it is the job of this committee to understand why. The scientific publishing ecosystem has changed dramatically. What was once a straightforward process of peer review and dissemination has become a complex, commercialized marketplace with maligned incentives and bad actors willing to exploit them. Now consider the scale of this problem. Estimates suggest over 400,000 published studies worldwide that may originate from so-called paper mills. Operations that produce fabricated and manipulated manuscripts for a fee. In 2023 alone, Wiley, a prominent American publisher, retracted over 8,000 fraudulent papers from a single subsidiary. This is not an anomaly. It's a symptom. The incentive structure driving this are well understood but underaddressed. Academic careers are built on publication counts. Institutions compete on research output. The result is a publish or perish culture that rewards quantity over quality and creates a ready market for shortcuts. When speed and quantity displace rigor or reproducibility, it is not just individual researchers who are harmed. It is the integrity of the science itself. Federal taxpayer dollars support the research that feeds this system. If fraudulent publications are being cited in grant applications that shape the direction of federal policy, American taxpayers will end up subsidizing corruption of the very science that they are paying to advance. Meanwhile, the Chinese Communist Party has built an academic incentive structure that has generated an industrial scale paper mill. Chinese universities have reportedly paid authors substantial cash rewards for publications in prestigious journals. A survey of medical residents at hospitals in Southwest China found that nearly half admitted, and that's just the people who admit, to buying papers, selling papers, or hiring ghostwriters to meet publication requirements. Critically, this fabricated research does not stay in China. It enters Western journals, shapes citations, and can contaminate the global scientific literature on which researchers depend. The rise of artificial intelligence intensifies all of this. AI lowers the cost of technical barriers to generate plausible looking scientific content at scale, even if it's inaccurate. The same tools that can accelerate legitimate research can be weaponized to fabricate data, manufacture citations, and flood the publication pipeline with content that appears credible but is not and is often referenced in other papers. As these capabilities grow, the gap between what looks like science and what actually science is will widen. Unless institutions, publishers, and federal agencies act deliberately to close that gap. Open access mandates, which this committee has engaged with extensively, add another layer of complexity. The movement to make federally funded research freely available is a worthy goal. But the article processing charge model used to make this financially feasible creates its own distortions. This model incentivizes publishers to prioritize quantity over quality, disadvantaging smaller institutions and enabling predatory journals that collect fees while providing no meaningful peer review. The agencies under this committee's jurisdiction have begun responding. Following letters sent by this committee earlier this year, agencies are developing awards conditions that can lead to suspension or termination of funding when integrity requirements are not met. That is a positive development and we want to understand how those efforts are progressing. But this agency action has limits and structural pressures faced by this industry require a more comprehensive response. This committee's responsibility to ensure that federal investments are science protected and that research enterprise maintains this credibility is required to function. And that American institutions are not disadvantaged by the competitors that treat scientific integrity as optional. That is the purpose of today's hearing. Our witnesses can help us understand both the depth and the problem of what meaningful solutions look like. To further support our understanding and discussion on this topic, I ask unanimous consent to include in this hearing record letters from the American Association of the Advancement of Science, the International Association of Scientific, Technical, and Medical Publishers, Wiley, Springer Nature, and the American Phytopathological Society, without objection, so ordered. Thank you and I look forward to a serious and substantive discussion. I now recognize Ranking Member, representative from Ohio, for opening statement.

Thank you, Mr. Chair, and good morning to everyone here and welcome to our witnesses here today. This hearing scope is ambitious, to say the least. Published research connects the public to scientific discovery and without publication the impact of researchers' findings would be lost. Publication of research in a prestigious journal comes with understood and understood stamp of approval and experts and non-experts alike respect this. Journals set ethical and qualitative standards for the research that is fit to be published and therefore they are a platform upon which science-based decision making is possible. From the personal all the way up to public policy impacting the lives of all Americans. So we have this sprawling hearing this morning, arguably any issue that touches science writ large impacts the state of publishing. Publishers are extremely aware of the crucial role they play and frankly many of them charge handsomely for it. This issue should be taken seriously. Instead, we have the president's budget issued two weeks ago, which places a government-wide prohibition on using federal dollars for so-called prohibitively high publishing and subscription fees. This is an issue in need of a scalpel and this decree is a sledgehammer. Cutting federal funds for publishing costs is only going to put a further strain on our universities and researchers who are already under attack from this administration. This same budget also requests a $2.7 billion cut to higher education programs and billions dollars more in research funding. Research institutions foot the bill for these publication costs, meaning the true impact on scientific publishing is likely even more catastrophic. Ultimately pulling the rug from underneath the publishing industry carelessly will be a disaster for research integrity. Publishing fees contribute to high quality peer review to ensure that research published is in reputable journals is held to a high scientific and ethical standard. It is disturbing, though not surprising, that this administration is being so flippant about research integrity when high quality research also often undermines its anti-science goals. The amount of junk science that this administration has used to base its public policy should put a shudder down everyone's spine. To be clear, I acknowledge that this is a nuanced issue and I support the experimentation with different business models to expand public access. However, I'm also skeptical of whether the fees journals are charging researchers and libraries alike are fair at all. Ohio has been a leader in negotiating with publishers to expand access to journals through OhioLINK, a state mandated and funded effort that pools the purchasing power of universities to get Ohioans the best deal. This model saves more than a billion dollars, that is billion with a B, in journal costs every year. Through OhioLINK and similar transformative agreements, Kent State has saved more than a million dollars in article publishing fees and the University of Akron has saved $600,000 since fiscal year 22. Especially as AI slop continues to pose a threat to research integrity across the board, publishers certainly play a critical role in protecting research integrity, but transparency around these fees is warranted. I think it's important to zoom out on the many issues listed in this hearing scope. The administration throws around the words gold standard science as a distraction as they cancel millions and millions of dollars worth of research that has passed rigorous merit review processes. They eliminate scientific functions such as the EPA's Office of Research and Development and move a fraction of the scientists under the office of the administrator where they can be more tightly controlled by political operatives who want to silence research that is inconvenience for its deregulatory agenda. Words including accessible, disparity, and underrepresented are plugged into keyword searches to deny proposals for grant funding, ensuring that already marginalized researchers and understudied populations remain understudied. So while this hearing on paper strives to address serious issues, I bristle at the against the implication that the solution to these enormous problems is primarily found in assessing the state of the publishing industry. We would be far better served by more focused conversational on the evolving business model of the publishing industry given the changes required to adapt to public access policies. Then we could have a dozen oversight hearings on the other issues the other issues in this hearing scope about the myriad of ways in which this administration is systemically obliterating the high standards to which American scientific enterprise has held itself for decades. I look forward to hearing the expert testimony and to treating this issue with the nuance and perspective it deserves. Thank you, Mr. Chair, and I yield back.

Thank you, Ranking Member Sykes. I now recognize Ranking Member of the full committee for a statement.

Thank you, Chairman McCormick and Ranking Member Sykes for holding this hearing and good morning to our panel of witnesses. As some of you know, scientific publishing and in particular open access is a topic I've cared about and studied for a long time. I've been a cosponsor of the Fair Access to Science and Technology Research or FASTR Act since it was first introduced in 2013. Prior to that in 2010, I cosponsored the Federal Research Public Access Act. Now both of these bills had the goal of accelerating free public access to the results of taxpayer funded research. Congress unfortunately failed to act on these bills. On the other hand, the White House Office of Science and Technology Policy under President Obama took a big leap forward in federal policy when Dr. John Holdren's 2013 memo called on agencies to implement a maximum 12 month embargo on all federally funded research papers. And in her 2022 memo, Dr. Nelson as acting OSTP director under the last president, carried the ball even further toward the finish line. Now we can and should have a healthy debate about how we get to the goal of immediate and free access to the results of federally funded research and I hope we don't disagree about that goal itself. I will remind anyone in the room who needs to hear this, this research was paid for by the public. The intellectual value is created by the scientists and students doing the research and the peer reviewers are other scientists who donate their time. To be clear, I also believe that publishers have long played a valuable role in the scientific communication ecosystem. In the 20th century, it could be argued they played an irreplaceable role. But this isn't 1925 or even 1995 anymore and we have to be open to new frameworks for communication and new business models. To the credit of many in the publishing community of their own accord or perhaps pushed along by open access advocates, a lot has changed already. In particular, I recognize the multiple ways in which scholarly societies add value to the scientific ecosystem from conferences, mentoring, and professional development for students and early career scientists to public outreach. Unfortunately, there are publishers who are doing a disservice to US taxpayers, US scientists, and even to other publishers who are trying to do the right thing. These publishers are rapidly losing the overpriced subscriptions they've long depended on for their massive profits. They are exploiting new ways to monetize the full research cycle, even information about scientists intended to be private. For example, a few years ago, Elsevier bought an academic technology company called Interfolio so that they can repackage and sell information about federally funded science and scientists back to federal agencies. And for all we know, given the absence of any transparency or guardrails, they may be selling the same proprietary information to others, including to China. I recognize that building new business models in the age of the internet and now AI isn't easy for 100 plus year old nonprofit organizations, and I'm ready and willing to partner with nonprofit scholarly societies struggling to find their footing in this new world. But we can't stand idly by as a select few for-profit companies continue to exploit US taxpayers, US scientists, and potentially put at risk US economic and national security to line their own pockets. This is, as the ranking member has said, an incredibly nuanced issue. Thoughtful advocates for open access have valuable recommendations for how open access can address the many challenges identified in this hearing's scope. To that end, I would like to ask unanimous consent to enter into the record a letter from SPARC that is very thoughtful about the issues and the nuances that we face, Mr. Chairman. Mr. Chairman?

Without objection.

This hearing is wide-ranging in scope, all fascinating and important topics. I hope in addition to the breadth we can go deep on at least some of these important topics because

Without objection.

This hearing is wide-ranging in scope, all fascinating and important topics. I hope in addition to the breadth, we can go deep on at least some of these important topics because our science enterprise depends on getting this right. And with that, Mr. Chairman, I yield back.

Thank you, Ranking Member Lofgren. I now recognize Ranking Member, sorry, I recognize the chairman of the full committee for a statement.

Thank you very much, Mr. Chairman, for holding this hearing on this very important topic. Science is one of America's greatest engines of innovation, economic growth, and national security. The driving force of that engine is our ecosystem of scientific publishing, made up of journals, academic institutions, federal agencies, and researchers themselves. This ecosystem is undergoing rapid transformation due to evolving funding models, artificial intelligence, and increasingly competitive academic incentives. While some of these changes benefit the production and dissemination of science, others threaten productivity, innovation, and public trust. Earlier this year, this committee sent letters to federal science agencies expressing concerns about the growth of paper mills and predatory journals. These entities exploit the pressure to publish or perish, which can flood the system with low-quality and even fraudulent research. When bad science gets published, it wastes taxpayers' money, misleads policymakers, and can even put public health at grave risk. But our agencies are not the only ones who must protect the scientific publishing ecosystem. Academic culture and institutions and the publishers of scientific journals also play major significant roles. One issue is that the publish-or-perish culture prioritizes quantity over quality, creating opportunities for predatory journals and paper mills to profit while eroding scientific credibility. Peer review, the cornerstone of scientific quality, is also under stress. Standards are inconsistent, transparency is limited, and conflicts of interest often go undisclosed. The result is research that may pass peer review but fails reproducibility tests or misguides subsequent studies. Reproducibility and data transparency must also remain critical. Research findings are meaningful only if independent teams can verify them. And yet, data sets are often inaccessible, replication studies are undervalued, and open access policies are inconsistently enforced. A system that rewards transparency, reproducibility, and responsible data sharing strengthens public trust and ensures that science serves the American people. Additionally, emerging technologies like artificial intelligence and machine learning can be powerful tools for discovery, but they also introduce profound risks. AI can accelerate research, analyze data, and detect errors, but it can also be misused to generate fraudulent papers or manipulate research results. Without clear oversight, these technologies can amplify the very problems that we're trying to solve. And that's why we have invited our witnesses today to discuss how the publishing industry is addressing these issues and to identify best practices in the publication of scientific research. This hearing is about risk mitigation and building trust in both the scientific community and the publishers who serve it. Some best practices relate to retractions and corrections. This has been an area of increasing scrutiny. We are pleased to have Retraction Watch with us today to discuss how retractions are a necessary part of the scientific process, as well as the damage that can occur when they are delayed, ignored, or not publicized. Science must serve the public interest, not just the publishing industry or academic career advancement. By taking these steps, we can preserve the integrity of American science, strengthen trust in research, and ensure that taxpayer dollars are being invested wisely. I look forward to hearing from our witnesses today. I want to thank you all for being here, and I hope today's discussion will be the first step toward policies that begin to reinforce rigorous, transparent, and trustworthy science. And with that, Mr. Chairman, I yield back.

Thank you, Mr. Chairman. Let me introduce our witnesses now. Our first witness today is Mr. Carl Maxwell, Senior Vice President for Public Policy for the Association of American Publishers. Thank you for being here. Our next witness is Ms. Kate Travis, Managing Editor for Retraction Watch. And our final is Dr. Jason Owen-Smith, Executive Director, Institute for Research on Innovation and Science, University of Michigan. I now recognize Mr. Maxwell for five minutes to present his testimony.

Chairman McCormick, Ranking Member Sykes, and members of the committee, thank you for this opportunity to testify. My name is Carl Maxwell, and I'm here representing the nation's leading scientific and medical publishers. The Association of American Publishers appreciates the committee's leadership and continued attention to the policies that shape the integrity, accessibility, and long-term strength of U.S. research. AAP represents over 80 professional and scientific publishers, including dozens of domestic scholarly societies and university presses. AAP members employ or directly contribute to over 200,000 U.S. jobs and publish the world's leading research findings. Thirteen years ago, my mother testified before this committee on addiction research on behalf of UT Austin. She believed in the peer review and publication process as the gold standard for delivering best-in-class science and was intensely proud of her work as an epidemiologist and author. I hope I will do her legacy proud today. Publishers take pride in their role in publishing peer-reviewed articles and managing the scientific record to ensure the best and most accurate research. To have this responsibility is an enormous honor. But communicating results is not the endpoint of discovery. Rather, it's a stopover in an ever-repeating circle of questioning, research, and reporting. No article is a definitive answer to a question, but a point of data added to the record for future generations. Every day, hypotheses are proposed, experimented, and reported. Some results become foundational, others retracted and discarded. In science, mistakes are a natural part of the process. In fact, mistakes are a feature, not a bug. It is only by learning from our errors that we move collective knowledge forward. But it all helps inform the journey of science. AAP believes the best method for ensuring quality and communicating science is through a vibrant, competitive, and diverse marketplace with the maximum author choice. Some brief facts about the professional scholarly publishing community. Over five million articles are published every year. STEM publishing is remarkably diverse. Nearly 50 percent of all articles are published outside the top five publishers. Commercial publishers operate in an unforgiving information marketplace, responsible to shareholders. No STEM commercial publisher had annual publishing revenues greater than $4 billion. By comparison, last year Google's parent company, Alphabet, enjoyed revenue of over $110 billion in just one quarter. Most nonprofit scientific societies rely on publishing revenue to fund operations, fellowships, conferences, and professional development. Businesses and clinicians across America use this high-quality content to speed up scientific discovery, whether it's finding new sources of energy, exploring outer space, or finding the next cure for a patient. Scholarly communication now finds itself at a crossroads. Research funding growth has led to an explosion of articles. AI creates both incredible opportunities and challenges, and nations such as China and India continue to rise as competitors and collaborators. Paper mills and predatory journals show signs of organized criminal networks. Curating and publishing scientific literature is expensive and requires substantial upfront cost in technology and integrity checks. Publishers invest their own money in the publication of each article. Federal agencies now require grant-funded authors to give free copies of the article to the public, removing subscriptions as the primary business model, a policy known as immediate open access. To provide free public access to articles while also ensuring article integrity, publishers are developing new business models. Looking forward, policymakers must weigh important and sometimes competing priorities: promoting a system that accelerates discovery and enables emerging uses such as artificial intelligence, ensuring timely and affordable public access to research results, and managing taxpayer costs and administrative burdens on researchers. At the same time, any sustainable system must safeguard rigor, quality, and trust in the scientific record. The risk, if these considerations are not carefully balanced, is a system where speed and access crowd out verification, leaving credible evidence-based research struggling to stand apart from unvetted or misleading claims. It is important we get this right because if parts of the publishing ecosystem fail, the downstream impacts will move beyond U.S. publishing and affect U.S. innovation capacity and the quality of science treatments upon which all Americans depend. I look forward to a robust dialogue with my fellow panelists and the committee on the future of publication. Thank you.

Thank you. I now recognize Ms. Travis for five minutes to present her testimony.

Chairman McCormick, Ranking Member Sykes, and other distinguished members of the subcommittee, thank you for inviting my testimony today. I am Kate Travis, managing editor of Retraction Watch, a news outlet that has been covering retractions, corrections, and other events in scholarly publishing since 2010. Our parent organization is the nonprofit, nonpartisan Center for Scientific Integrity. For well over a decade, we have been sounding the alarm, we and others have been sounding the alarm about the publish or perish system. It creates pressures that can lead to questionable research practices. Its reliance on overly simplistic metrics enables mechanisms used to game the system, like citation cartels and corrupted peer review processes. Publish or perish is what has allowed businesses like paper mills and predatory journals to flourish, and more recently, is leading to an explosion of AI-generated papers flooding journals. We report on these topics, all relevant to today's hearing, through the lens of retractions. A retraction occurs when a journal decides an article can no longer be relied upon because of unintended mistakes or misconduct. While retractions can be evidence of misconduct, the fact that they exist doesn't mean science is failing. It means the scientific literature is being cleaned up. Retractions mark as unreliable papers generated by AI, articles created for a business transaction rather than genuine research, and of course, papers containing honest error and those with evidence of misconduct. Retractions don't stop articles from getting cited, as we and others have shown, but the hope is to mark unreliable science before it wastes resources. Retraction rates have risen from about 0.02 percent of articles in the early 2000s to about 0.2 percent today, in large part because of greater scrutiny. Those numbers come from our Retraction Watch database, a publicly available and widely used resource with details on more than 64,000 retractions. The database shows that retractions have been rising over the last decade, with over 5,000 papers retracted in 2025. Last year, the U.S. accounted for about six percent of retractions, India about 18.5 percent, and China about 42 percent. When we look to see what percentage of a nation's articles end up retracted, the global average is about 0.1 percent. The U.S. and the UK are well below that at about 0.04 percent, and China is well above that at about 0.3 percent, with Ethiopia, Saudi Arabia, Iraq, and Pakistan as high or higher. Transparency is key to finding problems, and this is an area where science could use your help. Too often, the people checking claims are unable to access datasets, even those produced using public dollars. And likewise, those of us reporting on misconduct are unable to obtain materials related to investigations by the publicly funded watchdogs like the NSF's OIG or the U.S. Office of Research Integrity or the Office for Human Research Protections. At HHS, the ORI released just two misconduct findings last year, a drop from an average of 10 to 12. The NSF has lost key personnel charged with research investigations. We need more investigations and more transparency around them. The U.S. government has helped police scientific integrity in powerful ways, setting an international standard. The Justice Department has joined qui tam suits under the False Claims Act to prosecute cases of scientific fraud. To the extent whistleblowers are protected and rewarded, they are also empowered to stop bad actors. The Justice Department has also occasionally filed criminal indictments against scientists in cases where scientists have squandered public funds through fraud. An editorial in Science last week suggested establishing a national database of misconduct investigations to inform universities considering researcher candidates found guilty of misconduct elsewhere. The intersection of science and capitalism has afforded additional avenues for the government to prosecute offenses against the scientific process. In 2019, the FTC obtained a $50 million judgment against the OMICS International publishing group for engaging in predatory practices. This kind of action could be taken more often. We note that the SEC could exercise a degree of oversight over scholarly publishers whose companies are publicly traded. Focusing on publishers' bottom lines remains arguably an underutilized way to get their attention. But most urgently, as your invitation hinted, science needs a reform of the publish or perish system, which perversely encourages and even rewards mischief and misbehavior. By relying on overly simplistic numerical metrics to tell us who is doing the best science, we dangerously conflate quantity with quality, output with innovation, and publication numbers with reliable knowledge. To help the committee with this important work, we've provided additional details as references in my written testimony. Thank you again for the honor of presenting to you.

Thank you. I now recognize Dr. Owen-Smith for five minutes to present his testimony. Thank you. I now recognize Dr. Owen-Smith for five minutes to present his testimony.

Chairman McCormick, Ranking Members Lofgren and Sykes, distinguished committee members, thank you for inviting me to speak with you. I'm Jason Owen-Smith. I run a research data repository at the University of Michigan, where I've spent my career using data about research to understand how discovery works and to try to find ways to improve it. You have my written testimony. It's buttoned down and academic and it feels like it's about half footnotes. My kids would say it's very on brand. I'd be delighted to answer any questions you have, but I'd like to spend my five minutes today to tell you why the state of scholarly publishing is absolutely maddening to someone like me and what I think can be done about it. One of America's greatest superpowers is our capacity for discovery. That's fueled by publicly funded R&D. Publishing is a setting where American taxpayers fund research and American researchers do the work, but a few private, mostly foreign-owned companies end up with almost total control of both the results and data about how those results are used. The publishers profit handsomely from both those things. We seem poised to double down on that model, and the stakes are higher than ever. Here's what I mean. Big for-profit publishers aren't primarily in the publishing business anymore. They're sophisticated data analytics companies that use tools built around the scientific literature they publish to collect data on what researchers read, write, cite, submit, review, download, search, share, collaborate on, etc. Then they sell products built from that data to governments, funders, companies, and research organizations, our collaborators and our competitors all around the world. They're moving aggressively to expand their data about research and researchers. Let me make that concrete. Elsevier, the world's largest scientific publisher, which is based in the Netherlands, bought a tool called Interfolio in 2022. Hundreds of universities use Interfolio to manage academic hiring, annual reviews, tenure and promotion. Interfolio links unpublished documents like faculty research statements, grant proposals, working papers, even reference letters, the most sensitive kinds of pre-commercial, pre-competitive information there are in science, to private university HR data about researchers, and they use it for academic evaluation. Like thousands of other researchers, I use Interfolio all the time, which means Elsevier is likely tracking me. I'm an expert in exactly these kinds of data, and I can't tell you what Elsevier knows about me and my work, how they use it, or who can see it, which disturbs me a little. What I can tell you is that their privacy policy, which is online on their webpage, a credit to them, says they may use data they collect to create or improve products and they may share it with certain partners. A list of possible data sharing partners they link to on that website includes two Reed Elsevier IT offices in the People's Republic of China. If I proposed a data sharing arrangement with those parameters, sensitive private data, some of it likely involving critical technologies with collaborators in a CHIPS and Science Act designated country of concern, the compliance conversations that would follow with both my federal funders and my university would be, let's call them intense. Maybe those discussions happened at Elsevier, I genuinely don't know. But I think someone with oversight authority over science and technology probably should. The good news is we don't have to keep doing things the way we always have. Proven, U.S.-based, publicly accountable platforms for handling and responsibly working with data like these already exist. They could be developed and combined. President Trump's Restoring Gold Standard Science EO points in this direction, so does his Launching the Genesis Mission EO, as does bipartisan open data policy going back years. What my written testimony recommends and what I'm asking you for today is two things: a regulatory framework that treats data about federally funded research as a strategic national asset, and support for the accountable public infrastructure that can use that data responsibly in the public interest. This committee has an opportunity to advance gold standard science and forward ambitious initiatives like the Genesis Mission while protecting the security of America's public research ecosystem and the privacy of the talented scientists who power it. Please consider taking these steps to make data about research a first-class strategic asset for the American people instead of another maddening missed opportunity. Thank you.

I thank the witnesses for their testimony. I now recognize Member Webster for five minutes of questions.

Thank you, Mr. Chair. Appreciate it. Ms. Travis, U.S. research grants are having in some cases been awarded to individuals reliant on falsified research originating from CCP paper mills. These paper mills may have redirected research dollars away from in my area, I mean Max Planck, the Florida Institute, Moffitt Cancer Center, Lake Nona Life Science Incubator, and Shands. So what changes in the grant application and review process would make it harder for fraudulent or CCP-affiliated research to serve as the basis for American federal funding?

Thank you for the question. I think the best asset we have there are the systems in place to identify misconduct. And that's that can be the U.S. Office of Research Integrity, it can be NSF's OIG. Those those two offices in particular, their funding has been well, I don't know anything about their I don't want to speak to their funding, but their staffing has been low and they don't necessarily have the resources to investigate and to take a closer look at the research that is being used for these in those grant applications. But I think supporting the systems in place to identify misconduct is perhaps the best change that could be made there.

Thank you. Mr. Maxwell, quantum research is a pretty involves a lot of complex mathematics and simulations and other things. And it's hard for anybody to evaluate. So is there a more is that make the field more vulnerable to fabricated research slipping into our literature?

I was told there would be no math. You know, I think that one of one of AAP's members is IEEE, which is one of the leading publishers of artificial intelligence and computing research. I think it really kind of comes down to the journal and, you know, the integrity protections they put into place. You know, I I can't personally speak to quantum computing, you know, as a, you know, on behalf of AAP. I know this committee's spent a lot of time investigating that issue. I'm not sure it's any more fraught with with problematic publishing than any other particular field.

Thank you very much. Okay, Dr. Owen-Smith, what steps is the University of Michigan taking to mitigate the concerns of lacking peer review, the conflicts of interest, publish or perish incentives and so forth, and scientific misconduct? What kind of safeguards are there?

Thank you for the question. I'm not in a position to speak for the University of Michigan as an institution. I can tell you some of the ways that scholars like myself work on issues like this with our students and with our professional associations. Many of those have to do with data sharing, with replicability, particularly with ensuring that the underlying code for data analyses are available in sources like GitHub and other repositories to be used to replicate data. The key issue, however, is to understand and be able to systematically make sense of and find best practices for the research process, which is what data about research of the sort that we've been discussing can systematically be used for, both to find instances that are signals, leading indicators of when incentives are misaligned or when there may be bad actors at play, and to identify the signals of when good practices are being followed.

Thank you very much. I yield back.

The gentleman's time has expired and I recognize Ranking Member Sykes for five minutes of questions.

Thank you, Mr. Chair. I am in preparation for this hearing a little struck perhaps by the hypocrisy of this particular hearing, especially considering that the administration currently is doing it seems as everything that they can to undermine scientific integrity, promoting junk science, and undermining the very people who have led us to scientific bounds. I mean, I could go into the president's budget even further than this committee as we were just celebrating the Artemis II launch and meanwhile there is a significant cut in the NASA budget all in the same week. And so thinking about how we are navigating scientific publication, which is the foundation, bedrock of these advancements, I'm looking for where the solution is rather than just more attempts to undermine the science. And so for example, the federal prohibition on the subscription fees doesn't seem like something that most even open access advocates would support because it risks the entire peer review system and scientific and ethical standards that high quality journals enforce. And so Mr. Maxwell and then I'll go to you, Dr. Owen-Smith, if you have any comment on these declarations in the president's budget and what it does to the publishing community and the scientific basis of knowledge that we have come to rely upon for the advancements in American exceptionalism.

Sure. I think a ban on paying subscription fees, I think there are going to be and I suspect in this hearing we'll have reasonable conversations about publishing and the cost of publishing and associated with it. But banning particularly federally funded scientists and in particular federally employed scientists from access to current scientific results seems to be sort of self-defeating to me. I'm not sure what good it does to tell a NASA scientist or an Oak Ridge scientist, you know, go find it on a public repository, you know, you can't have a login to Science magazine or Nature magazine or something along those lines. There's a second aspect of that, which is the ban on publication charges, which is something the NIH is in the process of looking at right now. But federally researchers are required to give the agencies a copy of the post peer review article. And again, we can have conversations about how that gets paid for. But if in the end the author pays a fee to publish and then has to give that article to the agency and is not able to pay for that publication out of the grant, it's like a political science 101 definition of an unfunded mandate.

I'm sorry to interrupt you, only have two more minutes left and I want to hear from Dr. Owen-Smith. Thank you, I apologize.

Thank you for the question. I think you're hearing, as you said, the important difference between a sledgehammer and a scalpel. There are many important distinctions here that need to be understood with good evidence. We don't have the capacity to develop that evidence right now. One of the key things I think that is important in this policy is the distinction between the subscription fees and the publication charges. I think one of the things that could make something like this very useful would be not to ban subscription fees, but to require transparency about how they're set.

Thank you very much for that. And I know that we rely on a capitalistic society and profits are important, but also scientific integrity is important as well. So we'll keep that under advisement. I want to talk about integrity a bit more. You know, we've been dealing with this AI slop and AI has generated hallucination citations, which in these are just made up examples. And so the Make America Health Again Commission reported from the a report from the administration cited seven papers that did not even exist at all. And so I'm thinking about this in both of the hats that I have. I'm a lawyer. I am also a person who was a research assistant as I worked towards my Master's in Public Health degree and knowing how important citations are in both of those worlds, you can live and die. I mean, in law school they basically tell you if you mess up a citation your life your career's over. And I've heard of several lawyers filing briefings with bad citations as well. So I know that there is an answer to this and you all are the experts and you're really going to have to figure out how to do it in five seconds or less. So whoever wants to take it.

Maybe written comments for the record? Is that a fair answer since we're out of time? Maybe written comments for the record? Is that a fair answer since we're out of time?

I think that is a fair answer. I think that is a fair answer. One thing I was thinking of and if you can give me a couple seconds, Mr. Chair, is thinking of legal papers when something is bad law there are updated citations and the research documents show you that as you're looking through, we call it Shepardizing and such. So perhaps there's something like that, Ms. Travis, that could be something a part of the work that you do. Thank you. I yield back.

The gentlelady's time has expired. I now recognize Representative Begich for five minutes of questions.

Thank you, Mr. Chair. You know, when you pay for a product from a research perspective, it undermines the discovery of truth. And that really is the purpose of scientific research, to discover what the truth is. Academic researchers in the United States have become a sort of high priesthood telling the public what is true and what is not true. And they rely heavily on public dollars in order to fund that scientific research. So the publishing apparatus as a safeguard for determining what research is valid and should be presented to the public and what research is not valid, it's a very important gatekeeper for the public to ensure that the money that they're spending through their tax dollars is being spent accurately, it's being spent wisely, and it's not being spent to produce a predetermined outcome. We're seeing that a growing number of studies that have that have that have been a result of academic research are now being seen as unreproducible. And this flies in the face of the entire purpose of the scientific method. Science should be reproducible. And you're hearing now calls from some researchers who say, well, you know, maybe we should be able to publish papers that can't be reproduced. That seems ridiculous to me. And I'm curious, are publishing standards and peer review so eroded that journals can cannot validate or will not validate results prior to publication, that we have to determine that something is not reproducible after an article or research has been published by a journal? And I'll open this up to anyone who'd like to comment.

Just speaking on behalf of the publication process, you know, the peer review process sends out and checks, you know, is a check on the quality integrity of the article itself. Whether there are going to be reproducibilities or studies or not has to sort of follow on after the publication of the article. You know, the we've we're going to be talking about a lot of incentive structures today. There's not a lot of incentive within the academic world to do reproducibility studies. It's kind of a dead end like academic career unless you can find the funding to support it. And so I think if reproducibility is something that we're really, really interested in, candidly there has to be opportunities for young researchers as they work their way up through the academic system to say, hey, I'm going to do reproducibility studies on this this recent publication and track it down.

Just to follow on on that a little bit and something you said earlier, I think, you know, it is important to remember that as you said as well, a single research paper is a single data point in a in a in a larger picture. And any single scientific paper is just a chapter in a larger story of a research process. And, you know, I think you said, you know, we want to make mistakes in science. We want to we want to try things and we want to fail and we want to figure out what what to help us get to to the right answers. And so in terms of determining whether a story is or whether a a scientific paper is reproducible before publishing isn't quite practical. But ways that you can help ensure that are things like open access, open data sets, and even open peer review to understand how what critiques have been made on that research ahead of time.

And one way that you might think about creative ways to do this is if there are ways to track when papers become particularly important using data on scientific process, for instance, it is possible to set up relatively cheap competitions as a means to replicate things as they become more important or to test out extensions. That is a way that has been shown in some fields, computer science does it very well for instance, to drive replication and extension of particular findings as they become important parts of the chapter of the story of science that Ms. Travis is describing.

I appreciate that input. And I think that's very helpful as it informs the committee with respect to interventions that could be made by Congress to make sure that the results of research that are being published are truthful, reproducible. I love the idea about having some economic incentives in place to make sure that there is funding for reproducibility of results. I think it would go a long way to making sure that what's coming out of the research community is trustworthy for the American public. And I see that I'm out of time, with that I yield back.

The gentleman yields. Now I recognize Ranking Member Lofgren for five minutes of questions.

Thank you, Mr. Chairman. Dr. Owen-Smith, your testimony was really pretty eye-opening in terms of the data. Obviously open access is a benefit to the American people. I think it's also one that is threatening to publishers' bottom line, and it's actually interesting to hear that the publishers are evolving their business model to package and monetize data about science being conducted, not just the research results. I'm wondering, can you give us an idea about what the data security risks mean for our country in terms of foreign corporations or countries being able to access number one, and number two, do you have some ideas on what Congress could specifically do to protect against this?

Thank you for the question. I think they're exceptionally important. The short answer to your first question is no. I spent some time trying to determine for one publisher, Elsevier, what the general contents of their data stores would be from public sources and I was unable to. I think transparency is a good thing. I run a public data repository that works with sensitive data about science. Our data dictionaries, file and field guides are available on my website. In terms of the second question, again, I do not know. But I do know that from my written testimony, looking at public documents about the financials of some of these corporations, they are selling products based on these data across sectors and around the world. These are not limited to...

So we can let our imagination be our guide.

Indeed. And it is important also to note that we are buying these resources and using them to make policy and to make decisions and we don't necessarily know how the data were constructed.

You know, this has long been something that has bothered me. I mean essentially, without the data acquisition, you have a situation where the federal government helps the scientists do their research. Then they have to publish that research because their advancement, tenure and everything else depends on the publication of their research. Elsevier gives a small amount of money to the nonprofit societies, encouraging them to then become advocates of this system. And then volunteers actually, the scientists who are the peers, they don't get paid to do it. I mean they're... and then it's published and they sell back to the people who did the work the results of the work. Elsevier, the public reporting is that their adjusted operating profit for the last year that was reported was $1.17 billion, right? This is a Belgian corporation and I'm just wondering, and I've thought for a long time, how do you break the cycle? Because because the scientists need to have the publications for their tenure and... but it's a corrupt system and it needs to be broken. I mean I'm even thinking rather than do the peer review before the data is released, maybe what you need to require is that all the data be released and then the peer review occurs because that would help prevent fraud, it would help to prevent... and actually it might break the cycle in some ways. I think we need to completely rethink how this is done. I don't think there's a disagreement that's a partisan one in terms of wanting a science that is sound, that is not corrupt, that is replicable. All of that should be something we all agree on and I think we do. How we get there I think we need to completely think outside the box and how we break this cycle that is really limiting the advance of science and really deeply unfair to the American taxpayers. And with that, Mr. Chairman, I yield back.

The gentlelady yields. And now I recognize the Chairman, Dr. Babin, for five minutes of questions.

Just in time, thank you very much. Ms. Travis, last year Retraction Watch covered the case of the Dana-Farber Cancer Institute where researchers used images and data that were misrepresented and or duplicated in support of grant applications to NIH. Several high-level affiliates were named as possible perpetrators. Considering that this fraud was only caught after a whistleblower brought it to their attention, what level of the publishing ecosystem including federal agencies is most responsible or should bear the responsibility for catching such misconduct? And how frequently does Retraction Watch identify cases where U.S. federal grant funding was used to support research that was later retracted due to fraud or fabrication? If you can't remember those two, I'll be glad to ask the second one again.

Well I can answer the second one pretty quickly, which is we don't have numbers on that data. I mean the main way that we learn of misconduct happening in federal grant applications is through official reports through official investigation reports from universities and from bodies like the ORI and OIG. So that's when we learn, so I don't have statistics on that particular thing. To your first question about in particular about the Dana-Farber case and who is held responsible in there, that particular case called out the principal investigators of researchers as being responsible or as overseeing labs where misconduct occurred, even though they personally weren't responsible. And we see this a lot in investigation reports that even if you are not personally responsible for misconduct, the people in your lab, you're responsible for the people in your lab and you're responsible for the work that they produce. But the bigger answer is that it's actually that research integrity is everyone's.

Yes ma'am, absolutely. Thank you for that. And Mr. Maxwell, predatory journals often mimic the appearance of legitimate publications, making it difficult for researchers to distinguish between them. What is the industry doing to protect its brand and to help researchers identify credible venues and is that effort adequately funded?

There's a number of and I'm happy to get you to follow up with your staff with specific examples from each individual publisher. Industry-wide, the STM Integrity Hub is doing is sort of the kind of the overarching effort we make to sort of provide integrity support and help researchers identify predatory journals from real journals. I think within AAP members, there's not really an issue of predatory journals so much because they are quality high-standard journals themselves. I think there's a lot of investment that has to go into this issue to ensure the quality and I think where you often see I think for researchers that get duped into giving their publications to predatory journals is when they're looking for a low-cost solution for publication.

Thank you very much. And once again, Ms. Travis, are there journals or publishers that Retraction Watch considers to have meaningfully improved their integrity practices and what did they do differently? Are there any journals or publishers you can identify that have practices or policies that are harmful to research integrity? Let's go ahead and out them.

I'm going to focus on your first question, which is for publishers that have improved. It's a tricky question because we talk to a lot of publishers all the time for the stories that we produce. I would call out one in particular which is the Institute of Physics, IOP Publishing, which has decided to donate the funds of the author charges that it takes in on papers that it later retracts. It has decided to donate those fees to a charity that supports subscriptions in low and middle-income countries. So that's a particular good. But in terms of other publishers, there are other publishers who are working to improve their research integrity teams. I think we hear that and we but then we also at the same time see publications that shouldn't have been published in the first place and that's what those integrity teams should be preventing.

Yes ma'am. And Mr. Maxwell, last question. Publisher perish incentive structures are influenced by universities, not the federal government. To what extent can Congress actually influence those institutional norms and what leverage does federal funding provide?

If I can have maybe 20 seconds to answer, that's an excellent question and I really appreciate you asking it. It is unfortunate that impact factor has become one of the determinant issues in determining quality. I think that there are a wide range of issues to do much broader research assessment and I think that there are ways that the agencies can influence how they award grants and how they score grants that can help award researchers who are taking a more holistic approach to research and not just looking narrowly at journal impact factor.

Gotcha. I'm out of time so I yield back, Mr. Chairman. Thank you very much.

The Chairman yields. I now recognize Member Bonamici for five minutes of questions.

Well thank you to the Chair and Ranking Member and thank you so much to the witnesses for this conversation. I think I want to start by acknowledging what I hope everyone agrees with within this room is the importance of federal investment in scientific research and Dr. Owen-Smith, you talked about it as a strategic asset and we have to keep that in mind. And the, you know, open public access has been a critical part of that and critical part of U.S. leadership. We should be welcoming scrutiny, we should be prompting good questions. And obviously from the conversation, the testimony, what we know, we have a problem to solve. But I want to make sure and I hope all of us do want to make sure that the way we solve it is actually solving it and not exacerbating it. Right now the President's budget request is imposing these severe limits on covering publishing and research access costs with federal grant money. Independent federal research offices have been dismantled, sometimes subject to political review, compromising scientific integrity. And now this is a time of address, we have more data available but less resources and ways to address these problems. So I want to note that the executive order, the gold standard executive order, mentions concern about public trust in science. And I share that concern about declining public trust in science, but this administration is contributing to it, not solving it. In fact, the Union of Concerned Scientists identified 562 unique attacks on science between January of 2025 and March of 2026. And using the label, just simply labeling something as gold standard science is not going to restore public trust. In fact, under the executive order, it could be political appointees rather than scientific experts making decisions and those decisions might be based on outcomes over evidence. So that's extremely concerning and I want to ask you, Ms. Travis, what guardrails should Congress create to protect the integrity of science from political interference and misconduct? What guardrails would work in this context?

I think, um, as I said before, thank you for the question. Um, I think, uh, as I said before, uh, having the office, having established offices that are tasked with ensuring, uh, ensuring the integrity of research done on taxpayers' uh, using taxpayers' money is important. And, uh, the, and I think I talked about some of the other, um, mechanisms that the government has in terms of, in terms of making sure that the publishing industry is within, um, is within the bounds of, of legislation. Um, but I, just, yeah, I would say having, having those offices that are tasked with, with, um, scrutinizing, um, the integrity of, of research is important.

I appreciate that, but it's who's making the decision in the office that's critical. Mr. Maxwell, the President's budget for fiscal year 2027 would prohibit federal funds from being used for publishing or subscription fees. And I just want to follow up on the earlier questions about this. How would that policy change affect researchers and agencies that are trying to access the best available science?

Sure, I appreciate your follow-up question on that. As I mentioned, for researchers at federal facilities, it effectively cuts them off from their subscriptions, assuming it is implemented, and their access to online immediate publisher and all the publisher resources that come with it, which is often more than just the article. Publishers themselves develop a whole range of research products that go along with that. Along with that, it also means any grantee essentially has to pay out of pocket to publish if a publishing fee is associated with the publication.

How realistic is that, that they're going to pay out of pocket?

It is actually fairly realistic. It's not unusual to have researchers have to pay out of pocket if they don't have institutional funds to back up their publication. Particularly in the case of non-large research institutions, not for my Michigan colleague here, but if you're at a smaller school like a Michigan community college, you may not have access to resources to help you with publication.

Speaking of Michigan, Dr. Owen-Smith, the administrative data about research activity can reveal lots of things as you explained: funding trends, collaborations, emerging fields. How can the federal government better leverage research data to guide our science funding decisions without just relying on proprietary datasets?

Having a publicly accessible, community-based infrastructure to do this work that is based on something like the extension model that so successfully drove agricultural research through the 20th century would be a really nice way to think about this. Engaging with the scientific community in various locations around the country to establish priorities and feed them up into the policy process on an evidentiary basis would be a great start.

Thank you. As I yield back, Mr. Chairman, I certainly hope we can all work together on a solution that actually works to solve the problems that have been identified. I yield back.

The gentlelady yields. I now recognize Member Rivas for five minutes of questions.

Thank you, Mr. Chairman, for holding today's hearing. As an electrical engineer and former STEM educator, I created a nonprofit to encourage girls and women to get interested in STEM. And part of the reason why I founded this organization, as we all know, women have lagged in entering and advancing in STEM fields and I know from personal experience how hard that can be as a woman in these non-traditional fields. Women are often credited less than men and are underrepresented as authors in scientific literature. Rather than addressing these inequities in STEM, these disparities are further exacerbated by the Trump administration's policy decisions that are affecting the scientific community. More specifically, we must hold the administration accountable for the harm that they're causing not just to women researchers, but to our nation's research security and integrity through their cuts to science. In March of 2026, the Proceedings of the National Academy of Sciences of the United States of America published an article discussing that when the Trump administration unexpectedly terminated nearly 2,300 active NIH research grants and withdrew 2.45 billion, it was early career investigators and women that were most negatively impacted. The article stated that the 2025 terminations unevenly affected women investigators at key career stages, magnifying long-term consequences for the U.S. biomedical workforce. This is on top of the President's fiscal year 2027 budget, which proposes a cut of nearly 55 percent over fiscal year 2026 to the National Science Foundation. We can't compete globally if we're actively inhibiting our ability to conduct scientific research. By ensuring that women are getting the credit that they deserve, we will help to narrow this gender gap and by introducing girls to STEM careers and to ensure differing perspectives are discussed at the research level. Science benefits from a diverse group of voices. Dr. Owen-Smith, I've discussed the long-standing gender inequity in scientific research publishing coupled with the administration's cut to science. We know that the chaotic attacks to science funding from the Trump administration have affected female scientists more than male scientists. Disproportionately attacking female-led fields will deepen the inequalities that female scientists already face and we will lose their contributions to many parts of our scientific ecosystems and so many have been working over the years to increase that representation of women in these fields. In your expert opinion, what fields are most at risk when considering these attacks on female-led disciplines?

Thank you for the question. You're pointing out really essential things here. Most importantly, American science is American scientists. We want to protect them, we want to engage all of the talent the country has in doing this important work. One of the things we know is that female-dominated fields in science are facing greater challenges under these budgets. Areas such as public health are some of those. One of the things that your comments are pointing to is precisely the importance of being able to see these things in detail. One of the articles that you cited, women are credited less than men in science, actually uses data that was developed by my research organization. And the way they did that was by using administrative data that saw every person paid on a grant and compared the students who worked on the grant to the students who showed up on the author list. We saw a lot of female scientists who were PhD students or postdocs who were working on the project who didn't make the author list. I wasn't an author on that paper, we made these data available responsibly for research use. This is a concrete example of the ways that having data about the research process available for broad research use responsibly under the right kind of regulatory environment can really crack into these details to get at the policy questions in a broader way than simply asking what are the fields. It's what are the levers that we could use to ensure that all of America's talent is aimed at our core problems.

Thank you. Like you said, they're already being left out.

The gentlelady's time has expired. I now recognize Representative Issa for five minutes of questioning.

Dr. Smith, let me just ask a question follow up. The organization that didn't recognize women was funded in part by the federal government but was not a federal they were not federal employees, right?

That is correct. The organization, sorry, go ahead.

Go ahead.

The organization I run, the Institute for Research on Innovation and Science, is a data repository that collects information about the direct cost expenditures of federal grants.

No, I'm just trying to understand where the people that were denied authorship credit were not federal employees but grantees.

Yes.

So did you discover what was wrong in the cultures of those organizations that did that basically diminished the work of those people?

I was not an author on the paper that reported this. The authors on the paper, which was published in Nature about three years ago...

I'm only asking because this is one of the great questions. Academia is always lecturing us as though, you know, we're the people who don't get it and then this is a result of academia.

This was very much an academic problem. These are university laboratories and research teams that were not doing what I believe they should have.

Well, this committee certainly could go a long way toward mandating that any touch, feel, or participation be included and the Data Act supports a certainly an unlimited amount of names that could be included with some deference toward correctly stating exactly what their goal was about, but at least they'd be listed as authors. Mr. Maxwell, switching subjects, I'm concerned about two things that I hope you can help me with. One is the growth of the universal mining by China of our research data. You know, some of the previous questions were about concerns about the reductions or non-increases in funding, but isn't it true that effectively we fund China's research in addition to China funding China's research as a result of our open and transparent system?

Yeah, I think one of the the important things when we're talking about data, and then I'll answer your the question about articles as well. Data generated by taxpayer-funded research doesn't have IP protections, it's open to anyone who wants to use it. According to the Nelson memo, this is the year of implementation where all that data has to be made publicly available in a repository. So the the publishers never at any point hold the data in any sort of, you know, protective way. It always remains with the author, it's always available to anyone seeking that data from the author. When it comes to the so in that regard, the Chinese have access to any U.S. data funded by taxpayer research. The Chinese are now moving to a closed publication system, at least that's the reports we're starting to see.

Absolutely they have moved to it.

Right, and they would like to see most of their publications done in leading Chinese journals. And so their their information is now moving behind Chinese paywalls even as we're moving all of our information to immediate public access.

So that's a scenario obviously of concern. A second area of concern is and perhaps you you can go into some detail, but the growth of AI creeping into scientific research essentially takes the the real research and then that which is known by a computer and merges them. Does that concern you that it it is making scientific research more voluminous but not necessarily more informative?

Absolutely. Absolutely. The I would say the publishing industry right now is at the tip of the spear of AI slop even more than anyone is seeing in their, you know, their social media feeds. Trying to figure out how to sort through that is what every one of my publishers is talking about now. Behind closed doors, how you sort the real from the fake is is at the forefront of everyone's mind right now. And it candidly it it requires both throwing money at it, but I think also Congress needs to think about what research we use and what publications we use to train these models and making sure and from our perspective, we believe the licensed version of record is what AI should be trained on as opposed to whatever can be found in the public domain.

Thank you. And I'll just close with Dr. Smith one more time because you do look at all this information. In this day and age of AI, in this day and age of although that which is published is copyrighted and yet it is free, it's sort of a a dichotomy of what we think of as copyrighted. However, once copyrighted, plagiarism is still plagiarism. Do we have the ability or can we have the ability in in deference to Mr. Maxwell to in fact quickly determine whether or not material is original or not original and annotate it digitally so that people could begin to realize the the true source of something or the absence of a source? And quickly because my time's expired.

That's something that's beyond my area of expertise. I'd be happy to look into it and consult with a couple of colleagues and return to it.

Okay, perhaps anyone could answer for the record. Thank you, Mr. Chairman. I yield back.

The gentleman's time has expired. I now recognize Dr. Foster for five minutes of questioning.

Thank you. And thank you, Mr. Chairman. You know, this is not a new area. I remember back when I was in grad school many moons ago hearing stories about how coming out of a developing country, there was a whole alternate world of spectroscopy, of molecular spectroscopy, where there were papers were coming out describing the theory, the calculations, the apparatus, the analysis, and the results. Then it was just diverging with the rest of the world's knowledge of spectroscopy and eventually I think they the English sent over a professor and found that there was no apparatus from these papers that were being published. And this was, you know, in the first half of the last century. And, you know, obviously large language models and AI are the slop is just making things exponentially worse. You know, back in 2022 when I was chairing this committee, we had a comparable hearing. And back then we used we used the tools available then to sort of hide the fact hide obscure plagiarism. And we put Fermi's famous paper where he basically said here are the neutron absorption and generation cross sections that that allow a chain reaction to proceed, you know, one of the most famous papers in physics. And we put that through something designed to obscure the you know, obscure the plagiarism to generate a plausible similar paper. And then we put it through the state-of-the-art checkers and it said, nope, this is a wonderfully creative new paper that and with no plagiarism that we can detect. And so it's the cat and mouse is just going to just continue here. I do have some hope that maybe certain kinds of, you know, very obvious mistakes in in papers may be amenable to looking at with AI. That you're going to see when you see a retracted paper, you're going to be say, okay, let's get a big sample of retracted papers and train the the agents on those. Is do you is any of you aware of efforts along those lines?

I am aware of some that are working with text. The most effective appear to work with a combination of text and other data about use. So papers that come from paper mills tend to also cite other papers from the same authors and the same paper mills.

Yeah, you have these whole networks of of people citing each other's paper to get boost their ratings.

Exactly. Exactly. So many many data features together with textual features tend to do a better job of predicting, but I don't believe anyone has knocked the problem yet though maybe...

Well, yeah, I was going to say one of I think one of the challenges that's like a a deeper cut on this issue is attaching really quality metadata to the papers, ones that can really be, you know, text and data minable and and actually yield AI results has proven to be a challenge even for publishers. And and so I think that that that's an evolving area where everyone the more information you can find that AI can use, at least integrity AI can use to find mistakes in the papers or notice a lack thereof of that metadata would be really useful. And it's something that I know we're working on.

Yeah, and this is there's a lot of work that's very comparable in just commercial AI looking at how do you get rid of hallucinations. And, you know, the way you get rid of hallucinations is actually check the logic and and check the numbers behind any assertions that are and just, you know, I still use these things for coding just to keep my brain from rotting. And so I've just the the rate of increase of of capacity of these things is really impressive. And I so I urge I it'd be nice to actually throw some federal money towards that those sort of efforts so that we develop this capability. That's a separate issue here. About a year ago, the interim U.S. attorney for the District of Columbia began sending letters to journals around the country suggesting that they were partisan in scientific debates and insinuating that they were suppressing what they called opposing viewpoints. So at that time, I think there were at least four journals who publicly reported receiving similar letters. The New England Journal of Medicine, Journal of American Medical Association, Obstetrics and Gynecology, and a publication of the American College of Chest Physicians. I believe these letters, at least one of which was made public, really represented dangerous and intentional overreach by the federal government into scientific research. And it's particularly concerning because so much of our nation's research is federally funded. So Mr. Maxwell, did any of your members receive this or similar letters? And if they did, do they do you think they had an impact on the their publications that they chose to publish?

So I will get you a a formal answer on the number. The the ones you quoted are the ones that I remember. I the New England Journal for certain I remember and JAMA as well. They are the two I remember, but let me check back and see if anybody else received them. I'm not aware of anyone doing a direct response because as I recall his tenure at that position expired and they were not followed up on by his predecessor by his successor.

Yeah, but the do you get is there any way of knowing whether it sort of cast a pall over, you know, all journals knowing that they might be targeted this way? Yeah, so I appreciate if you come back with a for the record, that would be very informative. Thank you.

The gentleman's time has expired. I now recognize myself for five minutes of questions. I would say one of the things that has become evident in our current standards is that whoever controls what people believe controls the money and power of this enormous government that we have. With a $7 trillion budget and a $30 trillion GDP, it's enormous. In 2024, one of the world's top journals published a study predicting the large decrease in global economic output without major changes to public policy and to climate policy. This study prompted global and domestic policy changes with trillions of dollars invested, trillions. And widespread media coverage of climate concerns before it was retracted. Ms. Travis, this has sparked significant concern and distrust in peer review process. What steps should policymakers and federal agencies take to protect our constituencies from undue or even harmful policies based on flawed science?

Thank you for the question. I think the particular case that that you're talking about, the paper that was retracted was because one of I apologize, it's either 160 or 1,600 of the data sets looked at in that study needed to be adjusted. And the adjustment changed the conclusion and by standards for retraction...

Well, what can we do to change that?

The we can we all do a disservice to science by not talking about nuance, by not by by saying that this is the single study that is changes everything. That study is one of many that have studied the same thing and it should have been scrutinized for being an outlier, not taken and and looked at in in total. So the single best thing that we can do is to acknowledge that science is complex and that the research process is complex.

And that's why the peer review process is extremely important and I think we neglected in this case. Mr. Maxwell, on a daily basis I get on my phone and look for data online which a lot of times is inaccurate. How should we balance open science goals with the need to ensure that there is quality quality outlets possible for us and is there any way to do this at a cost measure that's not exorbitant for us as the government?

Sure, thank you for the question. I think candidly when we think about open science, open science works at two different levels. There is a level it works at for actual researchers and the way they interact with the research and then there is a level of the public's access to that science. I think that conflating the two often creates, you know, I think creates structural problems within the scientific research ecosystem. And candidly policymakers need to ask themselves is making the article immediately available the day of publication worth the cost you take from authors having to pay, the loss of readership quality controls, because as subscriptions go away, that means you lose the ability of the readers themselves to judge the quality of the journals themselves. And so I think there needs to be a much broader to Kate's point about nuance, it is not just simply as simple as the article itself is the be-all and end-all of research. It's really what we want where it is in the research ecosystem and how we want to fund that and how it informs discovery and innovation.

One of the things that I'm going to go a little off script here because I think this affect all of us in very real ways recently. I thought it was interesting I tried to stay very much away from the political aspect sort of the the partisanship of this, this research. But it was pointed out that this administration had done some things that they were they were claiming as nefarious. I would go back to when this really started. AI really came into effect last presidency. I mean, it really started to have its main influence. Research also was warped in the last administration when it came to when I was a doctor treating COVID patients and I was censored by people who had data that was non-peer reviewed, non-scientific, and censored me on things I knew to be true as a doctor because they were the government and they controlled the truth. Then they saw this again applied in climate control studies, in climate change. A Democrat, hard-carrying Democratic physicist named Koonin wrote a very good book called Unsettled on climate change. He was on Obama's climate control panel and stated how we had been misusing science over and over again to divert trillions of dollars to make lots of billionaires and lots of millionaires on this pseudoscience, not peer-reviewed. Excuse me, it was peer-reviewed but that wasn't included in the study, not included in the publishment that the government funded. And I'll go a step further, it was so mismanaged that we are funding school studies with millions and billions of dollars to get predetermined outcomes. And if you didn't get the right outcome for this climate change, we were not going to fund you the next time. That's manipulation by the government. And with that I yield. I thank the witnesses for their valuable testimony and the members for their questions. The record will remain open for 10 days for additional comments and written questions from members. This hearing is now adjourned.