The Five Ways That Facebook Uses Artificial Intelligence to Counter Terrorism

in news •  7 years ago 

image

London Bridge Terrorist Attack Aftermath | In the aftermath of the London Bridge and Borough Market, the previous night, the police were placed on a closed road, where they were killed and many others injured (Sunday's total). On Sunday 4th June 2017, in the south London Borough of Southwark, England. (Photo by Richard Baker / In Pictures via Getty Images)

On Thursday, Facebook made known its anti-terrorism policy, revealing its various ways of using artificial intelligence to combat terrorist activity.

Written by Monika Bicker, director of Facebook's international policy service, and Brian Fishman, the manager of the anti-terrorism policy of the social network, the post explains that Facebook finds the majority of content with a terrorist character before the user community n ' Take notice.

Facebook also said that while its current efforts are focused primarily on ISIS and Al Qaeda and their supporters, the company also seeks to use technology to combat other terrorist activities in the future, with more Of 150 employees focused on this research. The authors of the post added that monitoring Facebook is not obvious, with nearly two billion users in the world writing on the site in more than 80 different languages. The authors have reported five ways in which The social network tries to counter terrorism.

Facebook uses image matching, which allows to recognize (faces) and prevent the loading of terrorist propaganda items, previously recorded images or videos, onto the site. The social network also experiences linguistic understanding, a function using artificial intelligence to locate terrorist content through "text-based signals." This analysis proceeds by means of an algorithm that takes its first steps in the detection of similar publications, "reads the text. "The machine learning algorithms operate on a reactive circuit and improve over time." The site uses algorithms to identify terrorist groups, including pages linked to this activity, tracking newsgroups or User accounts, and investigating the involvement or non-involvement of related activities in the support of terrorism. "We use signals to locate an account that is connected to a large number of accounts that have been disabled due to terrorist activity, or to locate an account that shares attributes similar to a disabled account," the authors continued.
The determination of the actors of terrorism that repeat their misdeeds is another way for Facebook to act, although the text warns that these people will "continually evolve their methods." Facebook is now working on ways to investigate The terrorist activity in the family of Facebook applications, including WhatsApp and Instagram, actors of terrorism frequently using multiple platforms. In addition to artificial intelligence, the post also discusses how Facebook uses human expertise and partnerships to combat terrorist activity on the site. "In order to identify more quickly the sources of terrorist action and slow down the spread of such content online, we partnered with Microsoft, Twitter and YouTube six months ago to announce a shared data platform bringing together The sole basis for fingerprints of photos and videos, to target content produced by or supporting terrorist organizations, "reads one further. "Governments and intergovernmental agencies also have a key role to play in putting in place expertise that can not be independently developed by companies.

"We learned a lot about the propaganda mechanisms of Isis and Al Qaeda through reports from agencies in different countries. We also participated and benefited from the support efforts of the collaboration branch bringing together organizations such as the European Union Internet Forum, the international coalition against Daech and the British Office of the Interior. "

The post also explains that although Facebook does not decrypt encrypted messages, the site "provides all possible information in response to relevant police requests, in accordance with applicable law and policy." That the site has comments and reports, as well as the services of terrorism and security specialists, as well as a team to respond to death threats confirmed by urgent requests from the police. Facebook also said working to create a counter-discourse through partnership programs, including collaborations with the Institute for Strategic Dialogue, as well as with community and Non-governmental organizations.

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!
Sort Order:  

Yet another reason I despise Facebook. The definition of terrorism is so nebulous as to include nation-states conducting regular military operations overseas. In other words, literally anyone could be a terrorist. That this doesn't seem to give a large number of people weighty pause is appalling. I get it. If you use Facebook you're opening yourself up to their terms of service, but customer pressure could force them to scale back their operations significantly in this regard.

Hell, it could even push people toward a place like Steemit, where no such nonsense and privacy-invasive operations exist.

the only terrorist crime organization that hates your freedoms is the federal mafia

Such policies come to light bcos of terrorist activities which are on the increase...I dnt blame Facebook...if you dnt have anything to hide then no complaints. My people say if the weaver bird learns how to fly without perching so also will the hunter learn how to shoot without aiming...my only problem with Facebook is their greed as peoples contents has no value, glad steemit is here to fill that wide gap...upvoted @hicmaster