Lessons the Spam Wars: Facebook’s fight to Kill Fake News is just Starting

Today Facebook announced they are taking steps to reduce the influence of fake news on their service.

It’s a great start and I’m glad they are taking steps to fight fake news. It did make me wish they had taken some of the steps three months ago when it could have changed the course of history. (I’m not saying would have)

In the late 1990s email spam threatened to make email a waste land. Your inbox used to be filled with daily offers for fake products, porn sites, and attempts to steal your password. Today it hardly happens.

What can we learn from the successful battle against email spam?

It will take time

Spam first started to be a problem on email in the mid 1990s. I didn’t start Brightmail until 1998. The CAN-SPAM law didn’t pass until 2003 around the same time we estimated that spam email exceeded the volume of legitimate email. Over the years large email providers implemented more and more controls to limit spam. It was a hard fought battle and we were making progress mostly by focusing on authentication — being able to identify the source of spam and then shut it down. Then the bots arrived.

Bot networks, and their human-based cousins click farms, bypassed most of the the technology and laws that were in place because they looked like individual sources of email. The fundamental problem was that a medium that allows any-to-any connections is vulnerable to a coordinated attack of lots of disparate sources.

It’s a cat and mouse game and the only hope is to make the cat smarter and smarter.

There is no silver bullet

It took a combination of technology, social and legal innovation to stop email spam. It will take a similar full assault on fake news.

You have to make human judgement calls

No matter how good the algorithms get, there will always be a need for humans to evaluate the edge cases. Facebook, Google, Twitter, and other platforms are worried that they will be in a position to make the call on what is and isn’t real, guaranteeing to piss off a big part of their user base. My advice is suck it up and get a dedicated team and start developing clear rules of engagement. Figuring out what are legitimate sources should be their first job.