‘Publication retractions from India have skyrocketed since 2022’

Achal Agrawal founded the India Research Watch (IRW) after his observations of grave academic misconduct in India. Dr. Agrawal, now a freelance data scientist in Raipur, told scientific journal Nature how shocked he was when a student spoke casually of using software to publish his work. This amounts to plagiarism, Dr. Agrawal said, but the student insisted his work had passed the university’s plagiarism checks. Now a part of Nature’s 10 – a list of people “who shaped science in 2025” compiled by the journal – Dr. Agrawal spoke toThe Hindu about why he decided to quit his university job and devote his time to driving the discourse on research misconduct in India.

India ranks third, after China and the U.S., for publication retraction numbers. Is academic misconduct such as plagiarism not taken seriously in Indian academia?

India actually ranks second since 2022. The number of retractions from India has skyrocketed since 2022. Even the percentage of articles getting retracted has seen a sharp increase. It’s true that even the discovered cases for misconduct are not taken seriously. We know of cases with more than 30 retractions due to manipulation being felicitated and awarded in a premier Indian Institute.

University Grants Commission (UGC) guidelines stipulate very minimal penalty even if someone is found having plagiarised 60% of the content. Other forms of misconduct like data manipulation are not even mentioned in the UGC guidelines. Also, most countries have central ‘Research Integrity Offices’ that investigate and follow up on cases of misconduct. In India such an office does not exist and the investigations are left to the institutes, which have a vested interest in letting their researchers continue to use shortcuts.

In what ways is software used unethically to assist published work?

The biggest concern right now is the use of gen AI to generate fake data, papers and reviews. Before gen AI, a lot of people also used paraphrasing software to avoid getting caught for plagiarism. They would take existing papers, pass it through software, and then check if the plagiarism percentage was lower than the threshold. A lot of papers like that were then submitted, reviewed and accepted without anyone reading them at all. They contain hilarious artefacts of paraphrasing: “big data” becomes “colossal information” and “artificial intelligence becomes counterfeit consciousness”, for instance. A lot of image manipulation also happens using popular image editing softwares, but there are increasingly sophisticated tools to detect image manipulation.

Why did you feel it was necessary to quit your university job to catalyse this discourse?

It is difficult to do this work while working at a university, as it creates tonnes of conflicts of interest as well as pressures from the university itself. But I did not quit the university solely to do this work. I was also volunteering at a government school in Uttarakhand.

However, not being in the university system gave me the freedom to do this work without any conflict of interest or institutional pressures. I was lucky to be able to do a few freelancing projects too, as a data scientist, to help me with finances.

How easy is it to prove misconduct?

Most of the cases being caught right now are the really lazy researchers who do a shabby job at hiding the unethical artefacts in their work. The clever ones are way more difficult to prove, especially since it needs cooperation and coordination of several bodies in some cases.

On an average a retraction happens two years after publishing, showing how rigorous the process for a retraction is. A lot of problematic papers that are flagged by sleuths are still not retracted even after clear proof of issues with the paper. Plagiarism detection software as well as gen AI detection software are not reliable and can’t be considered proof. Also a lot of paraphrasing escapes detection.

India’s National Institutional Ranking Framework (NIRF) now pulls up institutions if several papers published by their researchers have been retracted. What more needs to be done?

Penalty for retractions is a necessary step to control this problem urgently. However, it is only treating the symptoms and ignoring the root cause, which in this case are the flawed metrics in NIRF itself, which has given rise to excessive focus on publication statistics. There is an urgent need to review NIRF to make it much more robust and transparent.

For example, we don’t know which college was given how much penalty this year, making the penalty imperceptible, and thus ineffective. Also, because of the excessive focus on research, education has really taken a backseat in higher education. Many professors cut corners while teaching to devote more time to research, as it is what is measured and rewarded.

Tell us more about the portal you have created where whistleblowers can report breaches anonymously.

A lot of people are only willing to talk about issues anonymously as they, legitimately, fear retribution for whistle-blowing. So we provide this feature on our portal where one is anonymous by design. The person has the option of providing an anonymous email in case they wish to follow up. We receive about 10 tips a day, but many of them are quite generic in nature asking to look at a certain profile, or are clearly driven by vendetta based on the language of complaint. Certain complaints are genuine, which we follow up and sometimes post about through our handle.

However, a ‘Research Integrity Office’ will be much better placed in dealing with these complaints as they will have the authority to actually pursue and do something about it.

You now face a lawsuit by private university…

Yes, a civil defamation case has been filed against IRW and is pending. Interim applications were disposed off on December 8 and a copy of the order is now awaited.