AI now being used to make social media safer and friendlier

The battle against Internet troll, bullies and cyber terrorists has turned in favor of social media applications as they start using artificial intelligence to make the Internet safer and friendlier for all ages.

Recently, Instagram announced that their comments sections are now being monitored and maintained by AI. Some people say that the feature had been active since last year October but given how no-one noticed, it was likely in a test run phase. The AI’s job is to filter out all offensive or spam like comments from any Instagram posts.

Before Instagram got this feature, Facebook revealed how they counter terrorism online using artificial intelligence. The brief outlined the following as their current efforts:

• Image matching: if an image with terrorist affiliations had been previously uploaded, the AI will detect it and prevent it from being re-uploaded again before it even reaches the servers. 
• Language understanding: this early stage feature of the AI takes a look at text then checks to see if is similar to text that has already been flagged for supporting terrorism. Over time, it will become better through machine learning and hopefully not flag your innocent text as supporting terrorism.
• Removing terrorist clusters: aims to disband groups or pages or accounts in support of terrorism by checking to see if they are connected to already banned accounts or are sharing content that has been removed for being connected to terrorism
• Recidivism: basically, they learn from previous behavior patterns of terrorists when creating fake accounts and ban accounts that follow that pattern or are being created by the same users.
• Cross-platform collaboration: for any artificial intelligence system to become better, it needs a large data set to learn from. Facebook will be working with other platforms like WhatsApp, Twitter, Instagram, Google, Microsoft and governments to share information on the behaviors of terrorists so that the AI will get better at detecting and preventing terrorism propaganda from spreading. 

When I heard about WhatsApp sharing data with Facebook in order to achieve this greater goal, I didn’t feel quite okay with that. That’s because WhatsApp has made a reputation of not sharing user data and being very private and secure. However, we all have to keep trusting that they will only use what they need and not misuse user data and violate privacy further under this good pretense.

As much as machines are great, there are situations that need a more human touch to make wise decisions. Google has seen this and have decided to add extra measure to ensure that people online will be safe especially when viewing video content. On YouTube, Google will be increasing the number of people in their Trusted Flagger programmer. This is a group of people who report violations on their terms of service and Google says that they are 90% accurate.

They also decided to ensure that some videos on YouTube that may just be religious out bursts will be not monetized and a warning will be displayed to the user before they can view the content even-though it is not violating any rules. This means that the content will probably be less viewed but still accessible allowing free speech but also showing a clear position that the company has on such content.

YouTube will also be increasing awareness on the matter through their Creators for Change program which promotes YouTube voices against hate and radicalization. This is an effort to change some people’s minds who may have been recruited without seeing this content. 

With the recent developments that have been happening around the world in terms of terrorist attacks, it’s good to see large tech companies coming together to counter terrorism by using the advanced tech they have at their disposal.