LinkedIn Blocked 21.6 Million Fake Accounts in Six Months

[engine]

LinkedIn has reported it's blocked 21.6 million fake accounts in the months January to June, 2019.
Wow, that's a lot that were blocked.
It goes on to say that 95% of those accounts didn't get past registration. Of course, that does leave 5% fake accounts. To cover for that, LinkedIn used AI and machine learning to take down 98% of the accounts through automated defenses, saying that the rest were captured through manual review.

[blog.linkedin.com...]

[Robert Charlton]

That's an amazing number of fakes. As a mod here, I'm strongly agreeing that the sooner one stops those fake accounts, the better. My guess is that the AI is triggered by too many coincidences... resembling some of what I assume Google is still doing.

Also, it's worth noting that it's a very well written piece, by Paul Rockwell (whom I'd never heard of by name), one of the LinkedIn startup advisors.

From the end of his blog post...

When we stop fake accounts, we start more chances for economic opportunity. We are committed to using every measure available to maintain your safety, allowing everyone to access economic opportunity while feeling supported and secure.

[engine]

Fake accounts are rife, everywhere, so LinkedIn is no different to FB, or twitter, for that matter. It might also be another reason G+ went away.
Fake accounts do nothing but pollute, and that's where tracking down the accounts is a good thing.

It's just quite surprising how many fakes LinkedIn found.

[Mark_A]

What would the creators of these fake accounts expect to get up to with them?

Seems a bit pointless to me but I expect I am missing something :-)

[engine]

What would the creators of these fake accounts expect to get up to with them?

Here's a couple of things:-
Many of them are looking for opportunities to steal identities from the network.
Other things include spamming their stories and their links to networks.

[Robert Charlton]

Many of them are looking for opportunities to steal identities from the network.

We see a lot of fake users posting here with obviously fake accounts.

Reverse image search may make fake LinkedIN profile photos more easy to detect. I remember a couple of years ago that many "doctors" from third word countries were models wearing white coats, taken from medical ads.

[RhinoFish]

The anonymous web is garbage.
Offer "certified" members the ability to use a credit card to pay a 1-time $1 fee to identify themselves as a real person.

[tangor]

Would "certified with credit card" actually pay $1 to have their credit information held by a third party that might get hacked at any given time? Inquiring minds and all that ...