Fake news is one of the greatest threats of today’s world. More than a threat to the legacy media, fake news deceives people, creates fear and panic and sows the seeds of a divisive and tribal society, open to extremist policies.

FNBL is a different approach to this issue. Instead of fighting fake news in their origin FNBL diminishes their visibility and limits their impact in all spheres of society.

the Fake News Black List database

The fight against fake news: meet FNBL

Given the nature of human language, I dare say automation via AI will never be an all-around solution for fake news identification, not in the current state of the art, that is. Some kind of human intervention will always be necessary. Content such as images or videos is a different matter, using file checksum, content detection, etc, fake news can be successfully identified and then flagged.

This concept is not only a proposal for fake news flagging just for Whatsapp, Facebook or Twitter, it’s a global initiative that can be used by any social media or content distribution network. It can be a tool for journalists around the world too.

 

The concept

The concept is similar from the decades old combat against spam, with a mixed global contribution from automation and human intervention.

Spam detection has many known mechanisms perfected over the decades and some of them can be used in the identification of fake news.

The most simple form of spam check can be made via DNSBLs (DNS Black Lists), a list of known IP addresses, domains and spammers. The same concept can be used against fake news.

A centralized database, let’s call it FNBL (Fake News Black List), with known domains, IP addresses and social media accounts can be used for fake news or fake content flagging.

Fake news are always replicated throughout the internet, so it can be flagged and checked with great degree of confidence. FNBL could include scores for all kinds of content. Images and videos are easier to check by file checksum, for instance.

Just like spam, all sources and content can have a reputation score. Content from nytimes.com will have a score of 100, globalresearch.ca will have a much, much lower reputation score, for instance.

This process can be somewhat automated with contribution from social media networks and all different types of contributors. All these IPs, domains, social media accounts will change over time and, just like spam, false positives will happen, but just like in any spam DNSBL, there are mechanisms for dispute.

AI has a role for fake news identification too. Events like the death of the pope, some scandal with the president or any kind of instant trendy post can be detected and red flags raised for instant human review.

For other content, posts from obscure sources, new social media accounts, etc, there will be need for some kind of human intervention, and mechanisms for content flagging can be implemented, just like in current spam blacklists. Again, false positives will happen, but there will be mechanisms for dispute.

Taking a page from Wikipedia, the concept of editors can be a solution for human fake news identification. These editors, existing fact checkers, selected and known players in the fight against fake news will have a somewhat direct access to update FNBL, the rest of the community will have means to contribute, but given the delicate nature of information all submissions must be checked by peers in the first contributions, but just like content in FNBL, all editors and contributors will have a reputation score too.

A period for open dispute will be maintained in case of new content with the possibility to rectify events of false positives, after that the content score will be closed and only editors with higher reputation scores can change it.

An open log will be available for tracking any and all changes to FNBL.

Any post in the social media, any link shared via real time chat apps can be checked against this database, just like any e-mail received right now is checked against DNSBLs, and flagged with the returned reputation score, maintaining high levels of user’s anonymity.

One user’s post, be it text, image or video, in principle, should not be checked, but if that same post, image or video becomes viral, that check must happen.

Links from external sources should always be checked against FNBL.

 

FNBL, what it is and what is not

FNBL can and should include all kinds of content score, from any country and in any language. The cheer size of it will be daunting, but a global effort from all the big players in the industry and the community can make it real.

It’s not a tool for censorship, religious or political opinion. It does not maintain any personal content or information whatsoever. It’s a tool for public awareness. In a fast pacing always connected world all it takes for sharing something is a touch with no time for cross reference. It’s a tool to maintain confidence in information the public get.

Maintaining FNBL open to community contribution will make it transparent. Involving the big players of the industry will make it work. The contribution of credible fact checkers around the world will keep it plural.

Open, fast and credible fact checking is what the public and, most importantly, democracy needs.

The fight against fake news: meet FNBL

Given the nature of human language, I dare say automation via AI will never be an all-around solution for fake news identification, not in the current state of the art, that is. Some kind of human intervention will always be necessary. Content such as images or videos is a different matter, using file checksum, content detection, etc, fake news can be successfully identified and then flagged.

This concept is not only a proposal for fake news flagging just for Whatsapp, Facebook or Twitter, it’s a global initiative that can be used by any social media or content distribution network. It can be a tool for journalists around the world too.

The concept

The concept is similar from the decades old combat against spam, with a mixed global contribution from automation and human intervention.

Spam detection has many known mechanisms perfected over the decades and some of them can be used in the identification of fake news.

The most simple form of spam check can be made via DNSBLs (DNS Black Lists), a list of known IP addresses, domains and spammers. The same concept can be used against fake news.

A centralized database, let’s call it FNBL (Fake News Black List), with known domains, IP addresses and social media accounts can be used for fake news or fake content flagging.

Fake news are always replicated throughout the internet, so it can be flagged and checked with great degree of confidence. FNBL could include scores for all kinds of content. Images and videos are easier to check by file checksum, for instance.

Just like spam, all sources and content can have a reputation score. Content from nytimes.com will have a score of 100, globalresearch.ca will have a much, much lower reputation score, for instance.

This process can be somewhat automated with contribution from social media networks and all different types of contributors. All these IPs, domains, social media accounts will change over time and, just like spam, false positives will happen, but just like in any spam DNSBL, there are mechanisms for dispute.

AI has a role for fake news identification too. Events like the death of the pope, some scandal with the president or any kind of instant trendy post can be detected and red flags raised for instant human review.

For other content, posts from obscure sources, new social media accounts, etc, there will be need for some kind of human intervention, and mechanisms for content flagging can be implemented, just like in current spam blacklists. Again, false positives will happen, but there will be mechanisms for dispute.

Taking a page from Wikipedia, the concept of editors can be a solution for human fake news identification. These editors, existing fact checkers, selected and known players in the fight against fake news will have a somewhat direct access to update FNBL, the rest of the community will have means to contribute, but given the delicate nature of information all submissions must be checked by peers in the first contributions, but just like content in FNBL, all editors and contributors will have a reputation score too.

A period for open dispute will be maintained in case of new content with the possibility to rectify events of false positives, after that the content score will be closed and only editors with higher reputation scores can change it.

An open log will be available for tracking any and all changes to FNBL.

Any post in the social media, any link shared via real time chat apps can be checked against this database, just like any e-mail received right now is checked against DNSBLs, and flagged with the returned reputation score, maintaining high levels of user’s anonymity.

One user’s post, be it text, image or video, in principle, should not be checked, but if that same post, image or video becomes viral, that check must happen.

Links from external sources should always be checked against FNBL.

FNBL, what it is and what is not

FNBL can and should include all kinds of content score, from any country and in any language. The cheer size of it will be daunting, but a global effort from all the big players in the industry and the community can make it real.

It’s not a tool for censorship, religious or political opinion. It does not maintain any personal content or information whatsoever. It’s a tool for public awareness. In a fast pacing always connected world all it takes for sharing something is a touch with no time for cross reference. It’s a tool to maintain confidence in information the public get.

Maintaining FNBL open to community contribution will make it transparent. Involving the big players of the industry will make it work. The contribution of credible fact checkers around the world will keep it plural.

Open, fast and credible fact checking is what the public and, most importantly, democracy needs.

Awareness

Knowledge that something exists, or understanding of a situation or subject at the present time based on information or experience

Awareness

Knowledge that something exists, or understanding of a situation or subject at the present time based on information or experience

we facts