In an attempt to block extreme, violent, hateful speech on the internet, Facebook and Google’s YouTube are quietly rolling out a system that automatically stems the spread of extremist views without completely eliminating it all together.

According to two sources familiar with the project who talked to Reuters, the technology was originally developed to squash videos that violated copyright law.

The program looks for “hashes,” which is like a fingerprint assigned to each video, and when multiples of the video are found the software automatically eliminates the copyrighted material. That same philosophy may now be applied to content deemed hate speech or propaganda by tech companies.

The videos identified as “extremist” would be prevented from being reposted elsewhere on the internet, but would not automatically destroy content that has not been watched before.

Reuters says it sources would not say how much human review goes into the process nor would they identify how a video is flagged as extremist in the first place, which is certainly the crux of the controversy with this software.

Whether it’s humans or robots making the decision, the tech companies are drawing a line in the sand as to what is free and open speech and what is “extremist speech,” which is concerning to any advocates of free speech.

“Its a little bit different than copyright or child pornography, where things are very clearly illegal,” Seamus Hughes, deputy director of George Washington University’s Program on Extremism, told Reuters. He added online extremist content exists on a spectrum and tech companies draw the line in different places.

Facebook CEO Mark Zuckerberg (R) hugs U.S. President Barack Obama during the 2016 Global Entrepeneurship Summit at Stanford University on June 24, 2016 in Stanford, California.

The program seems to be the result of pressure applied by the Obama administration and European leaders who held a call with tech companies including Alphabet’s YouTube, Twitter, Facebook, and CloudFlare back in April.

The call discussed the development of software by the Counter Extremism Project, which unveiled its system earlier this week and urged tech companies to adopt the system. Facebook and YouTube seem to be taking the matter into their own hands and a Twitter spokesperson said they are still considering whether or not to adopt the Counter Extremism Project software.