This browser is not actively supported anymore. For the best passle experience, we strongly recommend you upgrade your browser.

Social Media Links

| 5 minutes read

Recent Research Misconduct Concerns Warrant More Effective Diligence: Using AI Tools To Improve Research Integrity Processes

In just the first few weeks of 2024, two separate cases of allegations of research misconduct have been publicized. In both cases, the alleged misconduct was discovered by independent “data sleuths” using artificial intelligence (AI). The use of AI tools by data sleuths to uncover possible research misconduct is becoming more and more common as AI software can review and compare multiple images, data sets, or text in manuscripts in a fraction of the time it would take to manually review the same data. We recommend that institutions incorporate AI tools and software into their research integrity programs and research misconduct investigations given the potential for the AI software to review allegations in a much more efficient manner and given the lengthy time required to complete research misconduct investigations under the current federal regulations. These AI tools could decrease the time an investigation may take and, as a result, speed up the time to correct the scientific record when misconduct is found.

The first misconduct allegations publicized in 2024 were reported in a research integrity blog by biologist Sholto Davi.1 The blog questions images in 57 papers authored by four top researchers at the Dana Farber Cancer Institute (DFCI), including President and Chief Executive Officer Dr. Laurie Glimcher and Chief Operating Officer Dr. William Hahn. The allegations quickly made the news, with student newspaper The Harvard Crimson being the first to publish the story.2 

In response, DFCI’s Research Integrity Officer Dr. Barrett Rollins announced the request to retract six manuscripts. He also noted that 31 manuscripts are in the process of being corrected, and another manuscript is still under investigation.3 The remaining 19 were manuscripts in which DFCI researchers did not have primary responsibility for the alleged data errors. Dr. Rollins also noted that some of the manuscripts had previously been flagged for review by DFCI. Many of the manuscripts have had concerns raised on PubPeer, a website that allows researchers to post concerns about scientific manuscripts. 

The DCFI Communications Director stated in an email that none of the allegations raised by Dr. David involved clinical research and that cancer treatment was not impacted.4 However, as alluded to in Dr. David’s blog post, cancer research (and thus cancer treatment) is built upon the foundation of basic research. Tragically, if that research is fraudulent, enormous amounts of money and time can be wasted trying to build upon that inaccurate foundation and, worst-case scenario, could lead to clinical trials of treatments that are not effective and could actually be harmful.

The second case of allegations of research misconduct so far this year was raised by an independent scientist who serves as a scientific integrity consultant, Dr. Elizabeth Bik, and was posted in her blog on February 1, 2024.5 Dr. Bik stated that she received a tip about “sloppy and potentially fraudulent work” in a research group at Brigham and Women’s Hospital (BWH) that was alarming enough for her to look into even without a particular paper or practice questioned by the tipster. This case alleges that a Harvard Medical School neuroscientist, who is the vice chair for research in the Department of Neurosurgery at BWH, falsified data and plagiarized images in 29 manuscripts. The most serious allegations surround a 2022 manuscript that allegedly contains images plagiarized from seven unrelated papers by different scientists and from the websites of two vendors. These allegations were published by The Harvard Crimson6 and The Boston Herald7 on February 1, 2024, and quickly made the rounds of other news reports.

Representatives from both Harvard Medical School and BWH declined to comment other than assuring that the allegations would be taken seriously and reviewed in accordance with institutional policies and federal regulations. While this is expected and necessary, both research integrity professionals and scientists are becoming increasingly frustrated with the length of time required to complete a misconduct investigation in accordance with federal regulations. The Public Health Services (PHS) regulations on research misconduct state that institutions must “Pursue diligently all significant issues and leads discovered that are determined relevant to the investigation, including any evidence of additional instances of possible research misconduct, and continue the investigation to completion.”8 Historically the Office of Research Integrity has required institutions to continue an investigation until essentially all of the respondent researcher’s work has been reviewed, which takes an incredible amount of time and delays any correction to the scientific record. Hopefully, the pending update and revisions to the PHS regulations will address this concern and allow institutions to proceed with investigating allegations, making findings, and correcting the scientific record before pursuing other leads and reviewing all related research. 

As we noted above, Dr. David and Dr. Bik used AI tools to uncover the alleged misconduct in each of their cases. Dr. David stated that he used a combination of AI image analysis software and manual review to detect the manipulations in the DFCI manuscripts.9 Dr. Bik also used the AI image analysis software as well as reverse image searches on Google to detect the manipulations in the BWH manuscripts.10 

Given the potential of AI tools and software to review allegations and reveal misconduct in a timely and efficient manner, institutions should consider incorporating AI software into their research integrity programs and research misconduct investigations. This could be especially useful if the federal regulation requiring misconduct investigations to “pursue all leads” is not revised. The AI software could decrease the time an investigation may take, hopefully speeding up the time to correct the scientific record when misconduct is found. Once an institution has become comfortable using the AI software, they may consider reviewing PubPeer on a regular basis to ascertain whether there are concerns posted about any of their researchers and then use the AI software to look into those concerns. Institutions may even get to the point where they can become proactive about research integrity rather than reactive and use AI software to conduct random reviews of images and data to ensure that misconduct is not happening. 


1. Sholto David, “Dana-Farberications at Harvard University,” For Better Science (January 2, 2024) 

2. Veronica H. Paulus and Akshaya Ravi, “Dana-Farber Cancer Institute Researchers Accused of Manipulating Data,” The Harvard Crimson (January 12, 2024)

3. Veronica H. Paulus and Akshaya Ravi, “Dana-Farber to Retract 6 Papers, Correct 31 Following Data Manipulation Claims,” The Harvard Crimson (January 22, 2024)

4. Evan Bush, “Allegations of research misconduct roil Dana-Farber Cancer Institute,” NBC News Science News (January 26, 2024)

5. Elizabeth Bik, “Problems in Harvard Medical School studies include images taken from other researchers’ papers and vendor websites,” Science Integrity Digest (February 1, 2024) 

6. Veronica H. Paulus and Akshaya Ravi, “Top Harvard Medical School Neuroscientist Accused of Research Misconduct” The Harvard Crimson (February 1, 2024)

7. Rick Sobey, “Harvard professor accused of plagiarizing images, falsifying data: ‘What I found was a huge surprise’,” Boston Herald (February 1, 2024)

8. 42 CFR §93.310(h)

9. Veronica H. Paulus and Akshaya Ravi, “Dana-Farber Cancer Institute Researchers Accused of Manipulating Data,” The Harvard Crimson (January 12, 2024)

10. Veronica H. Paulus and Akshaya Ravi, “Top Harvard Medical School Neuroscientist Accused of Research Misconduct” The Harvard Crimson (February 1, 2024)

© Copyright 2024. The views expressed herein are those of the author(s) and not necessarily the views of Ankura Consulting Group, LLC., its management, its subsidiaries, its affiliates, or its other professionals. Ankura is not a law firm and cannot provide legal advice.


research, compliance, life sciences, article, risk & compliance, healthcare & life sciences, healthcare & life sci advisory, healthcare compliance, disputes

Let’s Connect

We solve problems by operating as one firm to deliver for our clients. Where others advise, we solve. Where others consult, we partner.

I’m interested in

I need help with