US president Donald Trump’s war with social media platform Twitter has become a war on censorship. Trump has threatened to shut down the platform after it fact checked his recent tweets. He is not supposed to be able to do that because Twitter is a legitimate business and if he did he would face legal challenges. Instead, what Trump did was sign an executive order that would target social media companies like Twitter along with Facebook and Google regarding the way they curate content.

You are either a publisher, like news media outlets or just a platform that allows users to publish their own content. In that case, the platform should not regulate the content based on their agenda but rather based on policy rules and regulations. This gets even more murky because of the definition of terms. Some will argue that Twitter has every right to fact check tweets, even though they are just opinions. Others will say that Twitter selectively chooses users who they want to censor or regulate.

The final executive order was released May 28, 2020 and it specifically deals with platform liability protections under Section 230 of the Communications Decency Act. Here is the part of Section 230 that concerns the topic:

“No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider” (47 U.S.C. § 230)

Now you can read up on this but it basically defines what a platform is and what they are allowed to do in accordance with the law. Fact checking can actually be a good thing for Twitter since it filters misinformation and other fake news that often circulates on the platform. However, is it even possible to police each and every tweet from its millions of users? Surely, you cannot have human workers do this 24 hours 7 days a week all through the year. The resources would not be available. AI is also not yet an expert in interpreting context from given text. It can detect hate speech based on words, but then it has to look at the context to determine whether it is hateful or not. It could just be meant as a joke, but now it depends on who interprets it.

Trump is going after the “good faith” requirement for removing “objectionable content” of Section 230. A platform is immune to lawsuits because they are just a platform, not the publisher of the content. If a user defames or threatens anyone on the platform, you cannot sue them. Instead you sue the publisher of the content, which is the person that defamed you. That is why you can sue newspaper outlets but not social media platforms. What Trump’s executive order can do is remove these protections from the platform. In other words, if you state that “mail-in voting leads to fraud”, the platform should not fact check that but allow other users to make remarks against it. If the platform does interfere with their user’s content or if the subject of the tweet is offended by it, then they can be subject to lawsuits.

This seems to be the problem with social media platforms. They have become so big, beyond what they are able to control. It has no solution or fix other than to explore ways of trying to regulate content. Tweets were more about opinion and not absolute facts. For example I can tweet “My cat is colored blue” because that is what you want to say. Now what if the platform attempts to fact check everything you say including the remark I just made? Would the platform ask for further proof by requesting a digitally signed photo of the blue cat? I am pointing out how ridiculous this example is. Instead, the platform should just allow everyone to post whatever they like, as long as it doesn’t violate the user agreement and policy of user conduct. Now the question becomes, who should do the fact checking?

Many influencers and public figures have complained about unfair treatment of their content. Some of them are conservative, so there are accusations that platforms like Twitter are censoring them based on their ideology. Perhaps a better solution is decentralizing social media in general. This will keep the platforms neutral and put the burden of fact checking on independent verifiers. Rather than function as the arbiter of truth, the platform is just a host for publishers who can post or tweet whatever they want. However, they will be held accountable to misinformation by a decentralized network of fact checkers who can rate the content. When it comes to censorship on these platforms, I do agree that certain content should not be allowed, especially those that promote indecent behavior and violence that can be dangerous to others.

Reddit uses an upvote/downvote system which puts the most voted content on the top of their list. This makes it easier for other users to see and it quickly trends across the Internet. Being downvoted is like a rejection of the content and its publisher. The problem with this is that it can be based on popularity, not facts, so even if the content published spreads misinformation users will upvote it because they like whoever is spreading the story or agree with the posting. This doesn’t prevent the spread of misinformation, but only helps it gain more attention. If Twitter were to use a similar system, then it can be troubling since there are users on that platform who can take advantage of their status to spread misinformation through upvotes. This is more the reason why a system for fact checking becomes more important.

Perhaps Trump’s action might be seen as an opportunity for more decentralized platforms to emerge. Social media, if you think about it, is actually a monopoly or even oligopoly of a few platform providers. Twitter does not really compete with another platform like itself that is as popular. Facebook doesn’t have any close competition either, with a user base that is close to 1.70 billion users (according to Statista). When these platforms are stripped of their privileges that have uncontested rights to filter, censor and control the content users post, the competition can take advantage. This could very well open up Internet freedom that is independent of the platform’s biases and political agendas. That means no one who works for the platform can tamper or manipulate anyone’s content. This allows users to put more trust in their platform of choice because it is not biased to any content they share so long as they follow proper policy guidelines (e.g no child porn, hate group, racist remarks, violence against others, etc.).

Decentralized social media sounds like a great idea. The problem is, what revenue model will they adapt to become a viable platform? The success of Twitter has to do with ad revenues and partnerships that allow third party connectivity to their platform using an API (Application Programming Interface). For a decentralized social media network, what sort of incentive will keep it running? Perhaps the answer to that would be open source software clients that run as peer-to-peer apps, but you still need a back end service to host the platform. This could be incentivized using a similar model as Twitter or Facebook, with a token. Another problem would be the UX/UI of these apps. The average user already finds using DApps (Decentralized Apps) for cryptocurrency quite a challenge. Think about how users will have to adjust from username/password to private key/public key systems. The user interface has to be designed as simple as possible with easy to use features. Decentralized platforms also do not have a technical support hotline to help them if they get locked out or lose their private key. If the platform uses a blockchain, it could be all the more troubling since content becomes immutable once it has been posted. While the platform cannot remove or prevent anyone from posting content, this violates GDPR (Chapter 3 – Rights of the Data Subject) rules that should allow users the right to delete or modify their own content.

If this is about freedom of speech and censorship resistance, then decentralization is the way to go. Liability cannot be put on the platform for any content since it is not under a central organization or company. Perhaps a better application would be to use decentralized networks for fact checking information, and incentivize the fact checkers for their efforts. If 230 were repealed, it could also lead to more censorship of content since the platforms will want to avoid lawsuits. With decentralization, let the platforms concentrate on the platform and leave the fact checking to sources outside the company. That should remove internal biases too, which removes central control. These systems can use a blockchain for timestamps and user verification to prove that the fact checker is authenticated on the platform and not a bot. As a proof of concept idea this would require verified fact checkers. One way to implement this would be to have professional journalists who do plenty of research and are connected to valid sources become verified by a non-biased committee of independent media outlets. When a certain post or tweet is flagged, the system can randomly select 3 or more fact checkers to verify the information. They will tag the tweet as fact or misinformation and for transparency will list the fact checking sources.

There is still no way to fully guarantee a system will be unbiased. That is the biggest challenge to all of this. Facebook has been using AI to filter content but it still has not reached that stage where it works perfectly. Certain posts may be erroneously identified as hate speech, even though the context could be meant in a humorous type of manner which most people do understand but machines do not. Given 10 fact checkers, how many would you assume to have no biases even if their credentials have been verified? The media seems to have a certain leaning that fits a certain narrative, so if the tweet came from a person who differs in thinking they could just tag it as misinformation. The system will then require valid sources for fact checkers always to make sure that they are not just making it up based on their beliefs. However this works out, if something that is more open to allowing users full control of their data and content emerges, it could lead the big social media giants to rethink about their platforms. We will see how the executive order will affect the outcome for social media if and when it takes effect.