New Plagiarism Allegations Rock Crypto Community

Last week, the crypto community found itself embroiled in another plagiarism controversy as two larger and better-known players in the field hurled allegations at one another.

The case began when Polygon Zero, the zero-knowledge subsidiary of Polygon, published a post accusing competitor Matter Labs of plagiarizing its code.

According to Polygon Zero, they are the developers of Plonky2 and Starky, a pair of level 2 zero-knowledge proof systems. Such systems verify that a certain statement is true or that a task has been completed and compute that off the Ethereum blockchain. These systems are crucial to developer efforts to both speed up the process and make it more scalable.

Polygon Zero alleges that Matter Labs recently released their own such system entitled Boojum, which contains a “substantial amount” of code from the Plonky2 library. While such a use isn’t necessarily unethical or illegal, since Polygon placed the library under an MIT/Apache license, Polygon claims that Matter Labs did not provide proper attribution. 

To bolster their case, they posted a series of photos that appear to show identical code and comments but without clear attribution in the files.

Polygon also took aim at claims made by Matter Labs, in particular one that claimed Boojum is 10x faster than Plonky2, noting that the relevant code is identical between the two implementations. They claimed that the benchmarks showing Boojum as faster were faulty and handicapped Plonky2 unfairly.

Matter Labs, for their part, has responded to the allegations. In a post published on X (formerly Twitter), the company’s CEO called attention to the first line of the main file in the module, which reads:

“NOTE: we take a Plonky2 implementation of non-vectorized field as the baseline, with extra modifications for compile-time evaluations. Even though we can not use “const trait” for now, one can use “_impl” const fn methods in non-generic contexts”

– Matter Labs Comment

He went on to explain that less than 5% of Boojum is derived from Plonky2 and that both are implementations of the Redshift, which was first introduced by Matter Labs 3 years before Polygon Zero announced Plonky2.

The story raises numerous questions, including what has either side done wrong? What legal repercussions could there be? And where does this fit in with the larger crypto community?

A Community Plagued with Plagiarism 

As we discussed back in 2020, the crypto community has had a long list of plagiarism scandals. 

The biggest was in 2018 when it was discovered that the whitepaper for the then-recently launched Tron (TRX) coin was heavily plagiarized from other papers. The scandal wiped away some $13.65 billion off the coin’s market cap in just two weeks.  

A survey by the Wall Street Journal at that time found that 271 out of 1450 digital coin offerings had “red flags”, with plagiarism being one of the most common issues. 

In what is the most public and long-running battle, there is an ongoing dispute over who is the author of the original Bitcoin whitepaper, with Craig Wright claiming to be the pseudonymous Satoshi Nakamoto, but providing little evidence to support his claims

Despite that, he has registered the copyright to the original whitepaper in the United States and filed litigation all over the world, including libel cases against those who challenge his claim.

But even with this backdrop, this new case feels different. Not only is it no longer the halcyon days of the late 2010s and early 2020s for crypto, but it’s a different kind of plagiarism issue, one that raises unique questions that are worth taking a look at.

A Different Kind of Case

Right off the bat, there’s no doubt that Matter Labs did copy Plonky2 code. Not only is it plainly obvious in the screenshots, but Matter Labs acknowledges this. While there are disagreements about how much and how important the code is, the copying definitely took place.

Though Polygon Zero accused Matter Labs of copying the code without attribution, it does appear that there was at least some attempt at attribution by Matter Labs. However, it also feels as if that attempt at attribution was wholly inadequate.

To highlight this, Polygon Zero points to Matter Labs’ post unveiling Boojum, which mentions Plonky2 but doesn’t make it clear that actual code was copied. Instead, the post only refers to constants and general inspiration as being what was taken.

And to the allegations that Matter Labs’ attribution was inadequate, I do have to agree. Both the Apache and MIT licenses require clear notices of reused code. It is unclear if those notices were included elsewhere, but at no point does Matter Labs make it clear exactly what code was copied from Plonky2.

However, once again, Matter Labs seems to acknowledge this, noting that they could have done better on when providing attribution, but also claims that it had “sufficiently attributed” the code in the first line of the library.

As such, Matter Labs feels as if Polygon Zero misrepresented the issue by saying that there was no attribution, when the dispute is over whether Matter Labs met citation standards and norms, not whether they attributed at all.

While the disagreement is understandable, and even raises some legitimate concerns, this doesn’t feel like the kind of disagreement that should lead to this big and this public of a blow up. It seems like something that could have been much more easily and more quietly resolved by the two companies simply working out the details.

Bottom Line

In the end, this story harms Matter Labs, Polygon and the crypto community at large. 

The initial post made this sound like a major breach of ethics, and it has been heavily reported as such. But, while there are issues here that need to be addressed, in particular citation standards when reusing open-source code, those are complicated and nuanced questions and don’t likely indicate an attempt of one company to screw the other.

Open-source software creates an environment where intense competitors can and routinely do use code written by each other. That, in turn, can and does lead to disagreements.

While I understand the temptation to put your competitor on blast for taking your code (and sometimes even feel it is appropriate), as more details have come out, it’s become clear that this has less to do with deliberate and malicious plagiarism and more to do with citation standards.

I agree completely that Matter Labs failed to cite the code adequately. It’s also possible that the citation they did provide was an attempt by them to do what they thought was the bare minimum to give credit to a competitor. However, going this route is basically the nuclear option and ensures that everyone involved, including the whole of the crypto community, suffers.

As I pointed out in 2020, crypto has a real struggle in building trust with the broader public. It doesn’t matter how incredible the technology is if only a handful of people have any confidence in it.

It would have been much better if this issue was resolved between the companies directly, rather than in the public space. As it sits today, this is another blemish not just on these two companies, but crypto as a whole.

In other industries, this would be resolved through emails and phone calls, not posts on social media. While the issues and questions are legitimate, there are better ways to address them than in the public sphere.

I hope that’s a lesson all companies in this space can learn from. 

Want to Reuse or Republish this Content?

If you want to feature this article in your site, classroom or elsewhere, just let us know! We usually grant permission within 24 hours.

Click Here to Get Permission for Free