Facebook and its founder Mark Zuckerberg have been under a lot of pressure recently to address the spread of fake news content on its news feed. While Zuckerberg had previously been reticent about actually suppressing such content, the social network is currently working on implementing screening procedures using artificial intelligence. This will be applied to both posted content and live content.

Right now, the social network’s executives are saying that Facebook actually has the technical capability to address the proliferation of fake news, The Wall Street Journal reports. All they would really have to do is to implement existing solutions via the creation of actionable policies. Based on Zuckerberg’s insistence for the social network to remain impartial, however, there might not be much chance of this happening.

There’s also the matter of the social media company’s policy team being unable to figure out how to integrate AI censoring capabilities that would match current practices. As the director of artificial intelligence at Facebook, Yann LeCun says, finding the balance to preserve freedom of speech while filtering fraudulent content is the challenge.

“What’s the trade-off between filtering and censorship?” LeCun asked. “Freedom of experience and decency? The technology either exists or can be developed. But then the question is how does it make sense to deploy it? And this isn't my department.”

Aside from fake news, it would seem that the AI will also be used to flag live stream videos that are considered offensive, Venture Beat reports. According to Facebook’s director of applied machine learning, Joaquin Candela, work is underway to provide a means of automatically tagging any live content that can be considered hurtful, racist, sexist, or derogatory in any way.

This is supposedly part of the social media company’s bid to monitor the content that goes through its servers. Then again, as with using AI to address fake news, there’s still the challenge involving censorship.