Guidance needed for using artificial intelligence to screen journal submissions for misconduct
Open accessAbstract
Journals and publishers are increasingly using artificial intelligence (AI) to screen submissions for potential misconduct, including plagiarism and data or image manipulation. While using AI can enhance the integrity of published manuscripts, it can also increase the risk of false/unsubstantiated allegations. Ambiguities related to journals’ and publishers’ responsibilities concerning fairness and transparency also raise ethical concerns. In this Topic Piece, we offer the following guidance: (1) All cases of suspected misconduct identified by AI tools should be carefully reviewed by humans to verify accuracy and ensure accountability; (2) Journals/publishers that use AI tools to detect misconduct should use only well-tested and reliable tools, remain vigilant concerning forms of misconduct that cannot be detected by these tools, and stay abreast of advancements in technology; (3) Journals/publishers should inform authors about irregularities identified by AI tools and give them a chance to respond before forwarding allegations to their institutions in accordance with Committee on Publication Ethics guidelines; (4) Journals/publishers that use AI tools to detect misconduct should screen all relevant submissions and not just random/purposefully selected submissions; and (5) Journals should inform authors about their definition of misconduct, their use of AI tools to detect misconduct, and their policies and procedures for responding to suspected cases of misconduct.
AI snapshot
Written from the paper’s full text.
TL;DR
Journals and publishers are increasingly using AI to screen submissions for misconduct, which can improve integrity but also risks false or unsubstantiated allegations and unclear responsibilities. The paper offers five guidance recommendations emphasizing human review, reliable tools, author notification and response, screening all relevant submissions, and transparency.
Key findings
- AI detection tools are not error-free: Turnitin has a reported 4% false positive rate, and Proofig has an estimated 5.8% false positive rate; false positives raise serious concerns because they may lead to false allegations against innocent researchers.
- Most misconduct allegations are unsubstantiated: a study of 3,561 PHS-funded research allegations from 1994–2011 found institutions dismissed 87.4%, and in 2022 the Office of Research Integrity dismissed 85.9% of allegations it received.
- In a study of manuscripts submitted to nine American Association of Cancer Research journals from January 2021 to May 2022, 63% of 195 images identified as duplications by Proofig were likely due to error rather than deliberate misbehavior.
- The paper recommends that all AI-flagged suspected misconduct be reviewed by humans; that journals use well-tested tools and stay vigilant about undetectable misconduct; that authors be informed and given a chance to respond before allegations go to institutions; that all relevant submissions be screened; and that policies, AI use, and definitions of misconduct be disclosed.
Methods
- This is a Topic Piece/commentary, not an empirical study; it presents ethical analysis and guidance rather than original data collection.
- It synthesizes cited literature, news reports, regulatory examples, and prior studies—including the Proofig/AACR screening study and misconduct allegation dismissal statistics—to support its arguments.
- The main analysis is normative and argumentative, culminating in five recommendations for journals and publishers using AI to screen submissions for misconduct.
Limitations
- The paper does not report a formal limitations section for its own analysis.
- It states limitations of AI detection tools: false positives and false negatives, inability to detect all forms of plagiarism or subtle image manipulation, and inability to prove that a respondent acted knowingly, intentionally, or recklessly.
- Evident from the text: the recommendations are not empirically tested, and the analysis relies on selected examples and cited literature rather than a systematic review.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
0 of 30 cited sources checked · 0 supported · 0 flagged
- Interpretationnone given
Using AI tools to screen journal submissions for plagiarism and data or image manipulation is likely to enhance the integrity of published research.
“While using AI can enhance the integrity of published manuscripts, it can also increase the risk of false/unsubstantiated allegations.”
- Also stated in: § Background · hedged
“While using AI tools to screen journal submissions for plagiarism and data or image manipulation is likely to enhance the integrity of published research, it also generates some ethical concerns.”
Quote in the paperNot a substantial claim - Also stated in: § Background · hedged
- Premisecitation
Proofig has an estimated false positive rate of 5.8%.
“Proofig, for example, has an estimated false positive rate of 5.8% (Evanko, 2022).”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Evanko, 2022
Not checked in the free run.
- Premisecitation
According to US federal policy, a finding of research misconduct requires proof by preponderance of evidence that the person acted intentionally, knowingly, or recklessly.
“For example, according to US federal policy, a finding of research misconduct (defined as fabrication or falsification of data or plagiarism) requires proof by preponderance of evidence that the person acted intentionally, knowingly, or recklessly (Office of Science and Technology Policy, 2000).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Office of Science and Technology Policy, 2000
Not checked in the free run.
- Premisecitation
Misconduct-detection programs are not error-free.
“Although misconduct-detection programs perform well, they are not error-free (Haritonova, 2022).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Haritonova, 2022
Not checked in the free run.
- Premisecitation
AI tools cannot prove misconduct because they cannot show that the respondent acted knowingly, intentionally, or recklessly.
“While AI tools can enhance our ability to detect deviations from accepted research practices, they cannot prove misconduct because they cannot show that the respondent acted knowingly, intentionally, or recklessly, which requires delving into the person’s state of mind (Shamoo and Resnik, 2022).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Shamoo and Resnik, 2022
Not checked in the free run.
- Premisereasoning
Distinguishing between misconduct and honest error can be a resource-intensive and time-consuming activity.
“Because researchers’ behavior can be negligent or incompetent without qualifying as misconduct, distinguishing between misconduct and honest error can be a resource-intensive and time-consuming activity.”
Quote in the paperSubstantial claim - Interpretationreasoning
The use of AI tools to screen submissions could likely lead to unsubstantiated allegations.
“The use of AI tools to screen submissions could likely lead to unsubstantiated allegations, that is, allegations that lack a sufficient factual basis.”
Quote in the paperNot a substantial claim - Premisecitation
Turnitin has a false positive rate of 4%.
“For example, Turnitin, a popular plagiarism detection tool used by many academic institutions, has a false positive rate of 4% (D’Agostino, 2023).”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites D’Agostino, 2023
Not checked in the free run.
- Interpretationreasoning
False positives from AI misconduct detection raise more serious concerns than false negatives because they may lead to false allegations of misconduct against innocent researchers.
“While both types of error can undermine efforts to detect scientific misconduct, false positives raise more serious concerns than false negatives because they may lead to false allegations of misconduct against innocent researchers.”
Quote in the paperSubstantial claim - Interpretationreasoning
Using AI tools to screen journal submissions can increase the risk of false or unsubstantiated allegations of misconduct.
“While using AI can enhance the integrity of published manuscripts, it can also increase the risk of false/unsubstantiated allegations.”
- Also stated in: § AI errors increase the likelihood of false or unsubstantiated allegations of misconduct · hedged
“The use of AI tools to screen submissions could likely lead to unsubstantiated allegations, that is, allegations that lack a sufficient factual basis.”
Quote in the paperSubstantial claim - Also stated in: § AI errors increase the likelihood of false or unsubstantiated allegations of misconduct · hedged
- Premisecitation
Most research misconduct regulations and policies distinguish between misconduct (which is intentional) and honest error (which is not intentional).
“Most research misconduct regulations and policies distinguish between misconduct (which is intentional) and honest error (which is not intentional) (Shamoo and Resnik, 2022).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Shamoo and Resnik, 2022
Not checked in the free run.
- Premisecitation
Unsubstantiated allegations of research misconduct can cause considerable emotional, financial, and reputational harm to respondents.
“Unsubstantiated allegations of research misconduct can cause considerable emotional, financial, and reputational harm to respondents.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Benderly, 2016
Not checked in the free run.
- Premisecitation
Detection tools have limitations that can impact efforts to detect plagiarism and image or data manipulation.
“Detection tools have limitations that can impact efforts to detect plagiarism and image or data manipulation.”
Quote in the paperSubstantial claim - Premisecitation
Defending oneself against a research misconduct allegation is an arduous and emotionally-taxing process that can last years.
“Defending oneself against a research misconduct allegation is an arduous and emotionally-taxing process that can last years.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Benderly, 2016
Not checked in the free run.
- Premisecitation
A recent study of manuscripts submitted to nine AACR journals between January 2021 and May 2022 found that the majority (63%) of the 195 images identified as duplications were likely due to error rather than deliberate misbehavior.
“Correspondence between the editors and authors showed that the majority (63%) of the 195 images identified as duplications were likely due to error rather than deliberate misbehavior (Evanko, 2022).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Evanko, 2022
Not checked in the free run.
- Interpretationcitation
Rejecting a paper without informing the institution does not prevent the author(s) from submitting elsewhere.
“In cases where misconduct appears to be highly likely, rejecting the paper without informing the institution is not sufficient because it does not prevent the author(s) from submitting elsewhere (Oransky and Marcus, 2016), nor does it inform institutions that support and fund authors about their malpractices.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Oransky and Marcus, 2016
Not checked in the free run.
- Premisecitation
A study of 3561 allegations of research misconduct involving PHS-funded research between 1994 and 2011 found that institutions dismissed 87.4% of allegations because they did not meet the definition of misconduct or there was insufficient evidence.
“A study of 3561 allegations of research misconduct involving Public Health Service (PHS)-funded research between 1994 and 2011 found that institutions dismissed 87.4% of allegations because they did not meet the definition of misconduct or there was insufficient evidence (Loikith and Bauchwitz, 2016).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Loikith and Bauchwitz, 2016
Not checked in the free run.
- Premisecitation
Most misconduct allegations turn out to be unsubstantiated.
“Although the media and research integrity experts tend to focus on high-profile cases involving substantiated allegations of research misconduct (such as the Hwang Woo-suk or Diederik Stapel cases), most misconduct allegations turn out to be unsubstantiated.”
Quote in the paperSubstantial claim0 of 2 sources checked- cites Loikith and Bauchwitz, 2016
Not checked in the free run.
- cites Office of Research Integrity, 2023
Not checked in the free run.
- Interpretationreasoning
Using AI tools to screen submissions may increase journals' and publishers' responsibilities related to reporting or investigating suspected misconduct.
“Because using AI tools to screen submissions increases the ability to detect misconduct, it may also increase journals’ and publishers’ responsibilities related to reporting or investigating suspected misconduct.”
Quote in the paperNot a substantial claim - Interpretationcitation
Rejecting a paper without informing the institution does not inform institutions that support and fund authors about their malpractices.
“In cases where misconduct appears to be highly likely, rejecting the paper without informing the institution is not sufficient because it does not prevent the author(s) from submitting elsewhere (Oransky and Marcus, 2016), nor does it inform institutions that support and fund authors about their malpractices.”
Quote in the paperSubstantial claim - Interpretationcitation
In the future, ill-intentioned researchers may take advantage of generative AI tools to create fraudulent images that are nearly impossible to distinguish from real images.
“In the future, ill-intentioned researchers may take advantage of generative AI tools to create fraudulent images that are nearly impossible to distinguish from real images (Gu et al., 2022; Kim et al., 2024).”
Quote in the paperNot a substantial claim0 of 2 sources checked- cites Gu et al., 2022
Not checked in the free run.
- cites Kim et al., 2024
Not checked in the free run.
- Premisecitation
False or unsubstantiated allegations can incur needless costs on involved institutions.
“Furthermore, false or unsubstantiated allegations can incur needless costs on involved institutions.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Michalek et al., 2010
Not checked in the free run.
- Interpretationcitation
It is likely that a high percentage of research misconduct allegations generated by AI tools will also turn out to be unsubstantiated.
“It is likely, therefore, that a high percentage of research misconduct allegations generated by AI tools will also turn out to be unsubstantiated.”
Quote in the paperSubstantial claim- Rests on an assumption: The proportion of research misconduct allegations that are unsubstantiated in traditional reports is a valid predictor of the proportion of AI-generated allegations that will be unsubstantiated.
- Premisecitation
One study estimated the costs of a single research misconduct investigation to be more than $500,000.
“One study estimated the costs of a single research misconduct investigation to be more than $500,000 (Michalek et al., 2010).”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Michalek et al., 2010
Not checked in the free run.
- Interpretationreasoning
Bias could occur in screening manuscripts for misconduct if AI tools are applied only to certain types of manuscripts or authors.
“It is also important to note that bias could occur in screening manuscripts for misconduct if AI tools are applied only to certain types of manuscripts or authors.”
Quote in the paperNot a substantial claim - Premisecitation
In 2022, the Office of Research Integrity dismissed 85.9% of the misconduct allegations it received because they did not meet the definition of misconduct or there was insufficient evidence.
“In 2022, the Office of Research Integrity, which oversees PHS-funded research, dismissed 85.9% of the misconduct allegations it received for the same reasons (Office of Research Integrity, 2023).”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Office of Research Integrity, 2023
Not checked in the free run.
- Premisecitation
Studies have shown that researchers tend to be biased against submissions from low- and middle-income countries and low-prestige institutions.
“Studies have shown, for example, that researchers tend to be biased against submissions from low- and middle-income countries and low-prestige institutions (Harris et al., 2017; Tavoletti et al., 2022).”
Quote in the paperNot a substantial claim0 of 2 sources checked- cites Harris et al., 2017
Not checked in the free run.
- cites Tavoletti et al., 2022
Not checked in the free run.
- May combine two claims. 5
- Classified as probably not a substantial claim. 8
May combine two claims.
PremisecitationAI image-manipulation detection programs are mostly accurate but can make mistakes.
“AI image-manipulation detection programs are also mostly accurate, but they can make mistakes (David, 2023).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites David, 2023
Not checked in the free run.
Probably not a substantial claim: may not pass S4 (based).
Premisenone givenUsing AI tools to screen submissions increases the ability to detect misconduct.
“Because using AI tools to screen submissions increases the ability to detect misconduct, it may also increase journals’ and publishers’ responsibilities related to reporting or investigating suspected misconduct.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing), S4 (based).
Interpretationnone givenSince AI detection tools are openly available, ill-intentioned researchers can try them out before submission to ensure that their manuscript passes certain tests so they can avoid being caught by AI tools.
“Furthermore, since AI detection tools are openly available, ill-intentioned researchers can try them out before submission to ensure that their manuscript passes certain tests so they can avoid being caught by AI tools.”
Quote in the paperNot a substantial claimMay combine two claims.
PremisecitationCurrent AI tools can detect plagiarism of words but not other types of plagiarism, such as misappropriation of data, ideas, methods, arguments, images, tables, or unique forms of expression.
“For example, current AI tools can detect plagiarism of words but not other types of plagiarism, such as misappropriation of data, ideas, methods, arguments, images, tables, or unique forms of expression (Hosseini and Gordijn, 2020).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Hosseini and Gordijn, 2020
Not checked in the free run.
Classified as probably not a substantial claim.
PremisecitationThe research integrity community have been using AI tools to identify manipulated images and papers produced by paper mills or AI ChatBots.
“Recently, the research integrity community have been using artificial intelligence (AI) tools, such as Papermill Alarm, Proofig, FigCheck, ImaCheck, and ImageTwin, to identify manipulated images and papers produced by paper mills or AI ChatBots (Else, 2022; Oza, 2023; Sanderson, 2024; STM Integrity Hub, n.d).”
Quote in the paperSubstantial claim0 of 3 sources checked- cites Else, 2022
Not checked in the free run.
- cites Oza, 2023
Not checked in the free run.
- cites Sanderson, 2024
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationIn March 2024, Wiley announced it would begin using a new AI-powered Papermill Detection service to identify papers produced by paper mills.
“In March 2024, Wiley announced it would begin using a new AI-powered Papermill Detection service to identify papers produced by paper mills (De Rose, 2024).1”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites De Rose, 2024
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe editors of the Science family of journals announced that they will use AI to screen all submitted images for problematic manipulations.
“The editors of the Science family of journals, for example, recently announced that they will use AI to screen all submitted images for problematic manipulations (Thorp, 2024).”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites Thorp, 2024
Not checked in the free run.
Classified as probably not a substantial claim.
InterpretationreasoningUsing AI tools to screen journal submissions for misconduct generates ethical concerns related to harm, fairness, and transparency.
“While using AI tools to screen journal submissions for misconduct is likely to enhance the integrity of published research, it also generates some ethical concerns related to harm, fairness, and transparency, which need to be addressed.”
Quote in the paperSubstantial claimMay combine two claims.
PremisecitationCurrent AI tools can detect relatively unsophisticated forms of image manipulation, such as image duplication or repositioning, but not more subtle forms of manipulation, such as alteration of microdetails of represented structures.
“Current AI tools can detect relatively unsophisticated forms of image manipulation, such as image duplication or repositioning but not more subtle forms of manipulation, such as alteration of microdetails of represented structures (Bik et al., 2016).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Bik et al., 2016
Not checked in the free run.
Classified as probably not a substantial claim.
PremisecitationJournals and publishers are beginning to take advantage of AI tools to screen submissions.
“Now, journals and publishers are beginning to take advantage of these tools (Conroy, 2023).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Conroy, 2023
Not checked in the free run.
May combine two claims.
PremisecitationRespondents who can afford to hire an attorney may spend tens of thousands of dollars in legal fees, and their reputations may still be damaged forever.
“Respondents who can afford to hire an attorney may spend tens of thousands of dollars in legal fees, and their reputations may still be damaged forever (Benderly, 2016).”
Quote in the paperSubstantial claim0 of 1 sources checked- cites Benderly, 2016
Not checked in the free run.
Classified as probably not a substantial claim. May combine two claims.
PremisereasoningIt is unclear what journals or publishers will do after detecting suspected misconduct in a submitted manuscript and whether or how they will share this information with authors or institutions.
“However, it is unclear what journals or publishers will do after detecting suspected misconduct in a submitted manuscript and whether or how they will share this information with authors or institutions.”
Quote in the paperSubstantial claim
The paper relies on this without support.
The proportion of research misconduct allegations that are unsubstantiated in traditional reports is a valid predictor of the proportion of AI-generated allegations that will be unsubstantiated.
“It is likely, therefore, that a high percentage of research misconduct allegations generated by AI tools will also turn out to be unsubstantiated.”
The paper relies on this without support.
Human review of AI-flagged cases can reliably verify accuracy and ensure accountability.
“All cases of suspected misconduct identified by AI tools should be carefully reviewed by humans to verify accuracy and ensure accountability”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Other AI systems can detect flaws with digital images and suspected plagiarism.
“Other AI systems can detect flaws with digital images and suspected plagiarism [81–83].”
From Autonomous artificial intelligence, scientific research, and human values (2026) · cites this paper as [83]
Many scientific journals have already begun using AI tools to assist with the editorial process.
“Many scientific journals have already begun using AI tools to assist with the editorial process.”
From Autonomous artificial intelligence, scientific research, and human values (2026) · cites this paper as [83]
An AI system could wrongfully accuse a scientist of inappropriate data or image manipulation.
“For example, an AI system could wrongfully accuse a scientist of inappropriate data or image manipulation [83].”
From Autonomous artificial intelligence, scientific research, and human values (2026) · cites this paper as [83]
Cites
Lund, B. D., Wang, T., Mannuru, N. R., Nie, B., Shimray, S., & Wang, Z. (2023) · Journal of the Association for Information Science and Technology · doi:10.1002/asi.24750 · Cited by 919 · cited 1× · see citing works
“For a review of research ethics issues related to using AI in scientific writing, see Hosseini et al. (2023), Lund et al. (2023), and Stahl and Eke (2024). as fabrication or falsification of data or plagiarism) requires proof by preponderance of evidence that the person acted intentionally,…”
The Ethics of ChatGPT-Exploring the Ethical Issues of an Emerging Technology
Stahl, B. C., & Eke, D. (2024) · International Journal of Information Management · doi:10.1016/j.ijinfomgt.2023.102700 · Cited by 476 · cited 1× · see citing works
“For a review of research ethics issues related to using AI in scientific writing, see Hosseini et al. (2023), Lund et al. (2023), and Stahl and Eke (2024). as fabrication or falsification of data or plagiarism) requires proof by preponderance of evidence that the person acted intentionally,…”
The Prevalence of Inappropriate Image Duplication in Biomedical Research Publications
Bik, E. M., Casadevall, A., & Fang, F. C. (2016) · mBio · doi:10.1128/mbio.00809-16 · Cited by 287 · cited 1× · see citing works
“Current AI tools can detect relatively unsophisticated forms of image manipulation, such as image duplication or repositioning but not more subtle forms of manipulation, such as alteration of microdetails of represented structures (Bik et al., 2016).”
The ethics of disclosing the use of artificial intelligence tools in writing scholarly manuscripts
Hosseini, M., Resnik, D. B., & Holmes, K. (2023) · Research Ethics · doi:10.1177/17470161231180449 · Cited by 253 · cited 1× · see citing works
“For a review of research ethics issues related to using AI in scientific writing, see Hosseini et al. (2023), Lund et al. (2023), and Stahl and Eke (2024). as fabrication or falsification of data or plagiarism) requires proof by preponderance of evidence that the person acted intentionally,…”
Self-citationResponsible Conduct of Research
Shamoo, A. E., & Resnik, D. B. (2022) · doi:10.1093/oso/9780197547090.001.0001 · Cited by 224 · see citing works
Self-citationHow ChatGPT and other AI tools could disrupt scientific publishing
Conroy, G. (2023) · Nature · doi:10.1038/d41586-023-03144-w · Cited by 121 · cited 1× · see citing works
“Now, journals and publishers are beginning to take advantage of these tools (Conroy, 2023).”
AI-enabled image fraud in scientific publications
Gu, J., Wang, X., Li, C., Zhao, J., Fu, W., Liang, G., & Qiu, J. (2022) · Patterns · doi:10.1016/j.patter.2022.100511 · Cited by 103 · cited 1× · see citing works
“In the future, ill-intentioned researchers may take advantage of generative AI tools to create fraudulent images that are nearly impossible to distinguish from real images (Gu et al., 2022; Kim et al., 2024).”
The Costs and Underappreciated Consequences of Research Misconduct: A Case Study
Michalek, A. M., Hutson, A. D., Wicher, C. P., & Trump, D. L. (2010) · PLoS Medicine · doi:10.1371/journal.pmed.1000318 · Cited by 93 · cited 1× · see citing works
“One study estimated the costs of a single research misconduct investigation to be more than $500,000 (Michalek et al., 2010).”
Measuring the bias against low-income country research: an Implicit Association Test
Harris, M. J., Macinko, J. A., Jiménez, G., & Mullachery, P. H. (2017) · Globalization and Health · doi:10.1186/s12992-017-0304-y · Cited by 91 · cited 1× · see citing works
“Studies have shown, for example, that researchers tend to be biased against submissions from low-and middle-income countries and low-prestige institutions (Harris et al., 2017; Tavoletti et al., 2022).”
(2022) · doi:10.24318/jtbg8ihl · Cited by 65 · see citing works
Kim, J. J. H., Um, R. S., Lee, J., & Ajilore, O. (2024) · AI and Ethics · doi:10.1007/s43681-024-00439-0 · Cited by 42 · cited 2× · see citing works
“In the future, ill-intentioned researchers may take advantage of generative AI tools to create fraudulent images that are nearly impossible to distinguish from real images (Gu et al., 2022; Kim et al., 2024).”
A comprehensive survey on image authentication for tamper detection with localization
Chennamma, H. R., & Madhushree, B. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-022-13312-1 · Cited by 40 · see citing works
Science’s fake-paper problem: high-profile effort will tackle paper mills
Sanderson, K. (2024) · Nature · doi:10.1038/d41586-024-00159-9 · Cited by 39 · cited 1× · see citing works
“…research integrity community have been using artificial intelligence (AI) tools, such as Papermill Alarm, Proofig, FigCheck, ImaCheck, and ImageTwin, to identify manipulated images and papers produced by paper mills or AI ChatBots (Else, 2022; Oza, 2023; Sanderson, 2024; STM Integrity Hub, n.d).”
A review of the literature on ethical issues related to scientific authorship
Hosseini, M., & Gordijn, B. (2020) · Accountability in Research · doi:10.1080/08989621.2020.1750957 · Cited by 38 · cited 1× · see citing works
“For example, current AI tools can detect plagiarism of words but not other types of plagiarism, such as misappropriation of data, ideas, methods, arguments, images, tables, or unique forms of expression (Hosseini and Gordijn, 2020).”
Self-citationNationality biases in peer evaluations: The country-of-origin effect in global virtual teams
Tavoletti, E., Stephens, R. D., Taras, V., & Dong, L. (2022) · International Business Review · doi:10.1016/j.ibusrev.2021.101969 · Cited by 36 · cited 1× · see citing works
“Studies have shown, for example, that researchers tend to be biased against submissions from low-and middle-income countries and low-prestige institutions (Harris et al., 2017; Tavoletti et al., 2022).”
Artifical intelligence (AI) in decision making
(2021) · doi:10.24318/9kvagrnj · Cited by 31 · see citing works
Kleinert, S., & Wager, E. (2012) · doi:10.24318/cope.2018.1.3 · Cited by 20 · see citing works
Thorp, H. H. (2024) · Science · doi:10.1126/science.adn7530 · Cited by 19 · cited 2× · see citing works
“The editors of the Science family of journals indicate that the handling editor “will probe further and take steps that could include rejecting the paper” (Thorp, 2024: 7), but they do not say much about their process and whether institutions will be informed.”
The Essential Need for Research Misconduct Allegation Audits
Loikith, L., & Bauchwitz, R. P. (2016) · Science and Engineering Ethics · doi:10.1007/s11948-016-9798-6 · Cited by 17 · cited 1× · see citing works
“…of 3561 allegations of research misconduct involving Public Health Service (PHS)-funded research between 1994 and 2011 found that institutions dismissed 87.4% of allegations because they did not meet the definition of misconduct or there was insufficient evidence (Loikith and Bauchwitz, 2016).”
InfluentialAI beats human sleuth at finding problematic images in research papers
Oza, A. (2023) · Nature · doi:10.1038/d41586-023-02920-y · Cited by 11 · cited 1× · see citing works
“…research integrity community have been using artificial intelligence (AI) tools, such as Papermill Alarm, Proofig, FigCheck, ImaCheck, and ImageTwin, to identify manipulated images and papers produced by paper mills or AI ChatBots (Else, 2022; Oza, 2023; Sanderson, 2024; STM Integrity Hub, n.d).”
‘Papermill alarm’ software flags potentially fake papers
Else, H. (2022) · Nature · doi:10.1038/d41586-022-02997-x · Cited by 7 · cited 2× · see citing works
“Recently, the research integrity community have been using artificial intelligence (AI) tools, such as Papermill Alarm, Proofig, FigCheck, ImaCheck, and ImageTwin, to identify manipulated images and papers produced by paper mills or AI ChatBots (Else, 2022; Oza, 2023; Sanderson, 2024; STM Integrity Hub, n.d).”
A Quantitative Study of Inappropriate Image Duplication in the Journal Toxicology ReportsPreprint
David, S. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.09.03.556099 · Cited by 7 · cited 1× · see citing works
“AI image-manipulation detection programs are also mostly accurate, but they can make mistakes (David, 2023).”
Can you rescue a damaged reputation?
Benderly, B. L. (2016) · Science · doi:10.1126/science.caredit.a1600165 · Cited by 1 · cited 2× · see citing works
“Respondents who can afford to hire an attorney may spend tens of thousands of dollars in legal fees, and their reputations may still be damaged forever (Benderly, 2016).”
Flagged first, then by citations
Research works printed but not linked: 3
- De Rose G (2024). Wiley announces pilot of new AI-powered Papermill Detection service. Available at: https://johnwiley2020news.q4web.com/press-releases/press-release-details/2024/Wiley-announces-pilot-of-new-AI-powered-Papermill-Detection-service/default.aspx (accessed 2 April 2024).
- Evanko D (2022) Use of an artificial intelligence–based tool for detecting image duplication prior to manuscript acceptance. Chicago IL: Peer Review Congress. Available at: https://peerreviewcongress.org/abstract/use-of-an-artificial-intelligence-based-tool-for-detecting-image-duplication-prior-to-manuscript-acceptance/ (accessed March 15 2024).
- Oransky I (2024) Up to one in seven submissions to hundreds of Wiley journals flagged by new paper mill tool. Retraction Watch. Available at: https://retractionwatch.com/2024/03/14/up-to-one-in-seven-of-submissions-to-hundreds-of-wiley-journals-show-signs-of-paper-mill-activity/ (accessed 14 April 2024).
Cited by
Sorted by citations
Suchikova, Y., Tsybuliak, N., Teixeira da Silva, J. A., & Nazarovets, S. (2026) · Accountability in Research · doi:10.1080/08989621.2025.2544331 · Cited by 70 · see citing works
“While valuable, this approach lacks granularity in capturing specific contributions and, as Hosseini and Resnik (2025) note, both ICMJE and CRediT frameworks raise concerns about how declared roles are verified.”
Publication ethics in the era of artificial intelligence
Kocak, Z. (2024) · Journal of Korean Medical Science · doi:10.3346/jkms.2024.39.e249 · Cited by 56 · see citing works
“31-33 However, journals and publishers should be aware of their limitations and that they are not infallible.”
BackgroundStrzelecki, A. (2024) · Learned Publishing · doi:10.1002/leap.1650 · Cited by 31 · see citing works
“Therefore, when it comes to the automatic identi fi cation of texts, the possibilities are currently limited (Hosseini & Resnik, 2024; Weber-Wulff et al., 2023).”
The rising danger of AI-generated images in nanomaterials science and what we can do about it
Davydiuk, N., Krieg, E., Gaitzsch, J., et al. (2025) · Nature Nanotechnology · doi:10.1038/s41565-025-02009-9 · Cited by 21 · see citing works
The ethics of erroneous AI-generated scientific figures
Skulmowski, A., & Engel-Hermann, P. (2025) · Ethics and Information Technology · doi:10.1007/s10676-025-09835-4 · Cited by 20 · see citing works
“Suggestions for AI-based misconduct detection are currently being worked on (Hosseini & Resnik, 2025) and should include a fair and balanced guide for assessing inaccurate AI-generated images.”
The PubPeer conundrum: administrative challenges in research misconduct proceedings
Caron, M. M., Lye, C. T., Bierer, B. E., & Barnes, M. (2025) · Accountability in Research · doi:10.1080/08989621.2024.2390007 · Cited by 14 · see citing works
Perspectives of Artificial Intelligence Use for In-House Ethics Checks of Journal Submissions
Alnaimat, F., AlSamhori, A. R. F., Hamdan, O., Seiil, B., & Qumar, A. B. (2025) · Journal of Korean Medical Science · doi:10.3346/jkms.2025.40.e170 · Cited by 15 · see citing works
Autonomous artificial intelligence, scientific research, and human values
Resnik, D. B., Hosseini, M., & Hauswald, R. (2026) · AI and Ethics · doi:10.1007/s43681-025-00908-0 · Cited by 15 · see citing works
“For example, an AI system could wrongfully accuse a scientist of inappropriate data or image manipulation [83].”
Self-citationA harm reduction approach to improving peer review by acknowledging its imperfections
Cooke, S. J., Young, N., Peiman, K. S., Roche, D. G., Clements, J. C., Kadykalo, A. N., Provencher, J. F., Raghavan, R., DeRosa, M. C., Lennox, R. J., et al. (2024) · FACETS · doi:10.1139/facets-2024-0102 · Cited by 8 · see citing works
“…(Price and Flach 2017), may be better and faster at detecting flaws in data analysis than humans, although it may be best suited for routine tasks such as verifying references (as is already done by some publishers; Schulz et al. 2022) and detection of image manipulation (Hosseini and Resnik 2024).”
Seeing Is Believing? Scientific Misconduct and the Detection of Problematic Images
Bik, E. M. (2026) · Anesthesia & Analgesia · doi:10.1213/ane.0000000000008033 · Cited by 3 · see citing works
Ethical Considerations for Generative AI in Social Science Research
Brown, B. A., & Heitner, K. L. (2024) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3693-8557-9.ch006 · Cited by 3 · see citing works
Reviewing manuscripts for scientific journals: recommendations for early career scientists
Forero, D. A., Glatt, S., & Oermann, M. (2025) · BMC Research Notes · doi:10.1186/s13104-024-07060-8 · Cited by 3 · see citing works
Suchikova, Y. (2025) · NanoEthics · doi:10.1007/s11569-025-00477-w · Cited by 2 · see citing works
Bains, V. K., & BHAWAL, U. K. (2025) · Journal of Healthcare Research and Education · doi:10.25259/jhre_4_2025 · Cited by 2 · see citing works
Generative Artificial Intelligence in Academic Research
Arbabi, M. A., Monib, W. K., Abubakari, M. S., & Shafik, W. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-4982-4.ch003 · Cited by 2 · see citing works
When AI Writes the Letters: Recognizing Synthetic Authorship Patterns in Medical Publishing
Lupon, E. L., & Micicoi, G. (2026) · Publications · doi:10.3390/publications14020021 · Cited by 2 · see citing works
“Early recognition of repetitive phrasing, exaggerated thematic drift, and stylistic uniformity may become as essential to editorial hygiene as rigor is to clinical practice (Prillaman, 2023; Hosseini & Resnik, 2025).”
Wang, X., & Zhang, X. (2025) · Publications · doi:10.3390/publications13040061 · Cited by 3 · see citing works
“…artificial intelligence (AI) tools have transformed scholarly publishing, as they are able to play a significant role in the preparation of manuscripts or even in the knowledge production processes related to their publication (Garcia, 2025; Mututa & Tomaselli, 2025; Resnik & Hosseini, 2025).”
Charbonneau, D. H., & Zhang, M. (2026) · Accountability in Research · doi:10.1080/08989621.2026.2632083 · Cited by 7 · see citing works
Why researchers hesitate to disclose AI use: development and validation of a multidimensional scale
Kulal, A. (2026) · Ethics & Behavior · doi:10.1080/10508422.2026.2661699 · Cited by 2 · see citing works
Maglaras, L. A., & Katsaros, D. (2024) · Computers · doi:10.3390/computers13110287 · Cited by 1 · see citing works
Maintaining research integrity when using AI in Research
Woodhams, J. H. (2025) · doi:10.37672/ukrio.2025.06.maintainingriwhenusingai · Cited by 1 · see citing works
Woodhams, J. H., Parry, J. M., & Scott, J. A. G. (2025) · doi:10.37672/ukrio.2025.06.embracingaiwithintegrity · Cited by 1 · see citing works
Wang, Y., Li, H., Du, Y., Zhang, P., Shi, S., Ma, Y., & Wang, S. (2025) · Disability and Rehabilitation Assistive Technology · doi:10.1080/17483107.2025.2561247
RetractedKnowledge of misconduct amid North-African post-graduate dental students: A cross sectional study
Kallala, R., Khemiss, M., Azzouzi, I., Gassara, Y., Touzi, S., Harzallah, B., & Saad, H. B. (2025) · F1000Research · doi:10.12688/f1000research.148694.4
Kallala, R., Khemiss, M., Azzouzi, I., Gassara, Y., Touzi, S., Harzallah, B., & Saad, H. B. (2026) · F1000Research · doi:10.12688/f1000research.148694.5
Governing Artificial Intelligence in the Evolving Academic Publishing Ecosystem
Karaarslan, E., & Aydın, Ö. (2026) · Journal of Metaverse · doi:10.57019/jmv.1899774
Impact of AI Integration on Downsizing Intentions in Entrepreneurial Enterprises
Deshpande, S., & Bhatt, V. (2026) · IEEE International Conference on Interdisciplinary Approaches in Technology and Management for Social Innovation (IATMSI) · doi:10.1109/iatmsi68868.2026.11465455
“Empirical research demonstrates the desire to compete in a business environment to be considered by entrepreneurs, primarily for its fun element rather than winning or accomplishments [21].”
InfluentialKallala, R., Khemiss, M., Azzouzi, I., et al. (2026) · F1000Research · doi:10.12688/f1000research.148694.6
Editorial: Research ethics and integrity in the artificial intelligence era
Chigwada, J. P., & Ngulube, P. (2026) · Frontiers in Research Metrics and Analytics · doi:10.3389/frma.2026.1931866 · Cited by 1 · see citing works
Alhelaly, Y., & Ashofteh, A. (2026) · IEEE Annual International Computer Software and Applications Conference (COMPSAC) · doi:10.1109/compsac69091.2026.00162
Angelier, V. (2026) · AI and Ethics · doi:10.1007/s43681-026-01376-w
“Research-ethics scholarship likewise documents the expansion of automated journal screening ahead of agreed standards governing its evidentiary use [ 19].”
Influential
- Authors
- Hosseini, M., & Resnik, D. B.
- Affiliations
- Northwestern University (US)
- National Institutes of Health (US)
- Corresponding author
- Hosseini, Mohammad
- Venue
- Research Ethics
- Publisher
- SAGE Publications
- Published
- 2024-05-11
- Volume
- 21
- Issue
- 1
- Pages
- 1-8
- ISSN
- 1747-0161 (ISSN-L)
- 2047-6094
- PMC
- PMC11731884
- OpenAlex
- W4396834410
- Semantic Scholar
- 143051ba917aee1357ce5adc72e87568580de444
- Cited by
- 33
- References
- 23
- Citations by year
- 2026: 15 · 2025: 12 · 2024: 6
- FWCI
- 13.63
- Type
- journal article
- Language
- en
- Update notices
- correction (2025-10-25)
- Open access
- gold
- doi.org
- Licence
- Locations
- Research Ethics (publishedVersion, cc-by-nc)
- PubMed (publishedVersion)
- PubMed Central (acceptedVersion, cc-by-nc)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- National Center for Advancing Translational Sciences (UL1TR001422)
- Topics
- Social Sciences › Safety Research › Academic integrity and plagiarism
- Social Sciences › Arts and Humanities › History and Philosophy of Science › Publishing and Scholarly Communication
- Social Sciences › Decision Sciences › Information Systems and Management › Academic Publishing and Open Access
- Keywords
- artificial intelligence, plagiarism detection, image forgery detection, data manipulation, journal editorial policy, accountability, transparency, Committee on Publication Ethics, research integrity
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-07.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-07.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
