The difference is WSJ doesn't have an option that says "only show this ad to white readers".

That is a difference, but in degree, not in kind.

You're declaring voluntary and involuntary segregation to be the same. I can read Black Enterprise or the WSJ. I can't read something deliberately hidden from me.

Facebook doesn't collect explicit race information, it's inferred from user behaviour. You can choose to do the sorts of things on Facebook that black people do, then advertisers will think you are black. Just like you can choose to read Black Enterprise, and advertisers will think you are black.

Exactly. You had better make sure that your Facebook account doesn't look "too black", "too old", "too jewish", "too democrat".

What possible relevance could there be between "doing the things black people do" and qualifications for any particular job?

TFA is about age discrimination though, and Facebook does explicitly collect age information.

Using dozens and possibly hundreds of pictures of the user is not inferred information in any meaningful since. They know your race, most likely with higher than 99% confidence on most people from the images alone. I'm sure it gets even more exact when they start doing text analysis, looking at your likes, and using the image data of your friends and family.

Facebook explicitly collects your birth date though, doesn't it? (Question is both rhetorical and actual.)

And gives you explicit warnings not to lie about it. And limits the total number of times you can change it (so you'd better hope you lied when you signed up 5 or 10 years ago).

Facebook is not real big on the 'ol anonymity.

In practice they can't really stop you from lying, but just telling people not to lie and making the question sound serious can often be enough to stop the average person. I suspect most people put in their correct age.

Odds are also pretty good your friends are going to mention your age at some point while wishing you Happy Birthday. If Facebook's data collection is smart enough it could decide to trust them over you. I kind of doubt that it is smart enough to do that right now, but it probably wouldn't be the hardest thing in the world to do if people started commonly lying.

Media has never been responsible of showing the user all the information, they have control over what and how they say it.

I also think that having such a hard-line reasoning doesnt accept the fact that criterias for discrimination are boundless, and that force will only be able to punish a very small subset.

Have to find an alternate way of dealing with changes of this sort than punishment and state threats.

Yes, there are boundless ways to discriminate. Legally, most of them are irrelevant. We've picked a few well-defined things like race, gender, age, religion, marital status, etc. and said "These things you can't discriminate on". If you want to not hire someone because they wore a blue shirt and you hate blue shirts, you can do that.

Further, these aren't punished by force. The remedy is a civil lawsuit, not a criminal complaint.

That's a good point.

In the same way the difference between a stabbing and a murder is one of degree, and 2 inches depth.

You've missed the boat. It'd be akin to White Enterprise magazine checking the race of buyers before a purchase can be made or the ad being seen. That's obviously unlawful.

The question is whether FB is immune from the CDA (depends

I don't know if the rules are different in employment, but yes, selective advertising in newspapers with different demographics is considered evidence of "steering"/discrimination in housing. (in California)

That's apples and oranges. This is a single platform, using a variety of techniques to "disappear" protected classes of people from employment -- it's akin to laundering discrimination. Including companies like Enterprise Rent-A-Car is telling, as their business model has always been built around targeting new college grads to be branch managers and retaining them for <5 years.

If you work in tech, you should be very much against this, as this is yet another example of bad behavior by Facebook that will hurt everyone. They are hiding behind the CDA to pretend that they are just passive conduits, and will ultimately push the government to tighten regulation and weaken non-Facebook sites that publish user generated content.

Do you think that's the same thing as is alleged in the submission?