Facebook DEC AI: Detection of Millions of Fake accounts by the integration of Artificial Intelligence


Although, Facebook had bad experience with AI in the past, when their efforts of developing self-sufficient integrated program become sour and acted against them. Horror runs down the Silicon Valley, when the news about their system get out that in the media that AI took control of the system and their scrap chat bots start developing its own language to communicate. They shut it down immediately to save themselves from eventually break down of entire system.

This time around, AI turned out to be productive for them and save them fortune by lowering the skilled labor cost. They develop machine learning framework Deep Entity Classification which enables them to detect millions of fake accounts in very short period of time and it is much more efficient. In the past, a report about Facebook employees aired which details that they are accepting bribes for letting fake accounts to stay on the network which seems quite hideous because people might use these accounts to do illicit activities online like cyber bullying and spreading the fake news.


According to the report presented by Facebook CEO Mark Zuckerberg, he boosted on fact that Facebook successfully removed 3.2 billion fake accounts between the month of April and September as compared to the manual evaluation which is quite appalling and the number is 1.5 billion last year. This all become possible by the launch of AI DEC which eventually results in reduction of abusive activities on the platform by 20%.

I observe this trend in the last few days, when I continuously tried to make fake accounts on Facebook but face disappointment. While in the past, I used to make fake accounts to play pranks on my friends and never got detected. But with advent of AI, it becomes impossible to run fake accounts on platform.    

Previous
Next Post »