Op-Ed: Science fraud is an expensive disaster damaging real science. It has to be stopped.

What’s happening is that anyone can now be credited with fake scientific credentials based on publications created by “paper mills”, referring to bogus research papers. Anyone can be credited as the owner of IP of any kind, real or imaginary, including patents, based on these papers.

All you have to do is pay the paper mills, and there are your non-existent “scientific credentials”. It looks good. You take credit for it, and you’re now an “authority” on your non-existent subject.

You genius, you.

You can use this fake documentation to position yourself to get a grant, sell fake products, get investors, and/or simply rip off people.

This isn’t so much new as a known issue becoming a serious problem. The science news often includes information about science fraud, notably from publications like Nature, which is a prestigious place for research papers.

Making money out of science fraud

Fraud is a crime. If you buy or sell anything and misrepresent the value of what you buy or sell, it’s fraud. If you price a single Dorito at $10 billion dollars, and apply for a loan based on that valuation, it’s fraud. You see how this works.

The paper mill geniuses producing fake research papers do quite nicely out of it. They can also do PhD theses, high school essays, and people pay per page.

You can also get blackmailed when you’re committed to your fraud by submitting your “paper”, and that’s not likely to be cheap, either.

 You’re that sort of genius.

The risks of committing science fraud are real enough for any dedicated masochist. If you’re an Ivy League idiot, your very expensive education, academic post, and/or any other values can be totally trashed. Suspension, expulsion, and perhaps civil law cases against you can result.

If you’re somehow stupid enough to commit science fraud as an actual scientist in a real research environment, you’re personally extinct, and rightly so. Imagine doing a fake research paper in a real research institution, and your 1+1 = 2342.13434 outcome is discovered by a peer group of real scientists. You’d be better off dead. Much better off.

The real damage done by fake science

The sheer loathing of fake science expressed by many real scientists is real enough for even a mass media commentator. The real scientists are spending real money and time and tortuous peer-review experiences on real research.

Then some moron includes easily detectable garbage in their work or about their work. This is the QAnon of science, the deranged fringe making a lot of noises because some people broadcast it worldwide. The fake science needs to be disproved. That takes time and money, too.

Add to this enchanting tableau the attacks on science by the world’s ignoramuses, particularly those in high public and corporate offices. Add the many attempts to discredit real science, mysteriously backed up by fraudulent data and information-spin. Any bot can do that.

The estimated cost of managing science fraud is €145 billion per year. No matter how much you adore your organized criminals, psychopaths and other invaluable social assets, that’s very expensive.

This fraud wastes grants and research budgets as well as time and effort that could be better used on real science. Finding and fixing the problems created by this costly, useless drivel is also expensive. There are regulatory issues, too, for falsified data, compromised research contracts, and other fun stuff.

Zero tolerance is too much. We need to shut this down at the source.

AI generates paper mill stuff, but it’s also good and getting better at finding it. Most people who’ve examined AI work know there are patterns. There are LLM characteristics. There are artefacts, common usage, and “Copyscape fodder” where it’s easy to identify unrelated texts that mystically repeat themselves verbatim.

You can do that manually, too, but it’s a lot of work, particularly when looking at utter garbage. AI is the most appropriate, too, with oversight. This is specialized work, but it’s doable and should be able to spot the garbage on sight.

What’s missing is a real, scary disincentive.

A paper mill could be some mindless bozo with a laptop or an international network. The mindset is “habitual criminal”. That’s the problem. These people typically aren’t scared of the criminal law process.  They’re scared of being unable to make money. It’s a serious weakness.

Further up the scale, the criminal process works better. The very average white-collar criminal who enters into contracts based on fraud can be vivisected by law. Nobody wants to know them. Their networks fail. The risk is extremely high. They can go from high-flyers to lower-tier sewer rats overnight. They become instant nobodies.

What’s needed is some tough legal definitions:

Contracts enforcing data integrity.

Direct personal liabilities for anyone involved in scientific fraud.

Strong case law to back it up and make it stick.

RICO-based law for systemic fraud. After all, it is a racket.

Recovery methods for all money spent on fake papers.

This nauseating situation has to go.