Facebook plans to use U.S. mail to verify IDs of election ad buyers
reuters.com
reuters.com
This honestly seems like it solves literally nothing. I'd be shocked to hear that foreign powers didn't already have e.g. diplomatic staff living in the US who were buying ads and listing their home or non-consular work addresses...
Actually, is there any indication that listing a consular address is prohibited? Because if not, then not only does this seem to not solve anything, it would in fact fail with certainty to solve anything at all, except maybe give the USPS some extra much-needed revenue.
Seriously, what am I missing here?
Yes. I think I've heard an egregious story in late 2016 but can't find a reference.
For something less concrete, see RT's complaint against Twitter [1].
And a few other equally less-than-concrete report regarding Facebook[2] and others [3].
[1] https://www.theguardian.com/media/2017/oct/27/russias-rt-rev...
[2] https://www.theguardian.com/technology/2017/sep/06/facebook-...
[3] https://www.recode.net/2017/10/6/16419388/facebook-google-tw...
https://about.usps.com/forms/ps1583.pdf
And of course it is possible that Facebook simply won't post letters to them (there are only a few of them around).
Scenario:
A: open credit or debit card account (can be stolen ID)
B: choose e-statements
C: when card arrives, reshipper just relays the card data (number/expiration/code) via email or text
Nothing physical need be shipped overseas. As long as the account stays paid up, the genuine ID holder won't know without getting and carefully reading his credit report.
There's zero chance they're interested in the $2 million ad buy from Russian propaganda agents.
Facebook absolutely wants to solve this problem. It harms their brand, they are attracting scores of negative media and worst of all every day of inaction gets them closer to media-style regulation.
And they are unquestionably making changes from News Feed algorithm updates to more rigid rules around buying ads to additional factor authentication in this case.
They want the news story to go away, but they also want to stuff it all down Putin's throat, and Trump's as well. Russian trolls did not make up even a small percentage of their ad profits last year, I wouldn't assume they're protecting them in the name of a very small amount of money.
Which means that if they were caught working on behalf of the Russians whilst on US soil then they would be looking at decades of jail time. Also at least there would be a physical, independent paper tail back to specific individuals.
IMHO Facebook should just ban all political advertising but baring that this isn't actually a bad idea.
And since Facebook is likely to be restricting numbers/types of ads per account this could definitely make it prohibitive for a wide spread assault on Facebook ad system.
There's ways to get a few fake addresses, but it gets more complicated if you want thousands. If you run spend significant amounts of money for campaigns and run through just few accounts your activities are easy to detect.
Source: Russia’s troll identities were more sophisticated than anyone thought => https://www.theverge.com/2018/2/16/17021684/facebook-twitter...
The UPS Store chain used to be called "Mailboxes, Etc."
https://arstechnica.com/tech-policy/2018/02/mueller-flips-am...
One of the most glaring problems is dynamic ad pricing. Ads from different candidates can see wildly different prices based on how 'engaging' their ads are. In my mind this gives a clear advantage to incendiary, sensational, populist candidates. In some cases the difference in pricing between ads can be as much as 10X, based on how 'engaging' each ad is.
That's fundamentally unfair for someone running a clean, policy-driven campaign.
As an Australian i just find this refusal to do this as bonkers considering we have a much lesser issue when it comes to election fraud.
"In Australia, where voting is compulsory for all adult citizens, no form of ID is required to cast a ballot at an election; instead, voters are asked three questions before being issued a ballot, so that they can be checked off the electoral roll: (1) what is your full name; (2) where do you live; and (3) have you [voted in this election]"
- Wikipedia
I showed my ID at a voting poll on the Central Coast NSW in the 2013 federal election and in the previous state election of WA when i voted i was asked to show a drivers license.
That said there is zero evidence that illegal immigration correlates with increased election fraud.
A lack of evidence about correlation doesn't mean it can't be true.
Also your claim necessarily cannot be true unless there has never been a case of an illegal immigrant illegally voting, which there are.
Because now I can say that with 0 illegal immigrants, you will have 0 illegal voters.
This is kind of what i'm saying, that in a system where there is no ID checks, what stops fraud from occurring other than people stating "we haven't found it at a large scale at this time".
I live in Australia, why would i be interested in suppressing minority voters in the USA? I don't even get the chain of logic in claiming this. How does asking for ID when voting suppress minorities? Are minority citizen entitled to ID?
The reason the type of voter fraud you are suggesting is rare is because it's not that effective at swinging an election and it's cheaper to buy a politician's vote legally in the USA.
https://www.nytimes.com/2014/06/11/upshot/vote-fraud-is-rare...
* https://www.brennancenter.org/analysis/debunking-voter-fraud...
* https://www.nytimes.com/2016/12/18/us/voter-fraud.html
* https://www.washingtonpost.com/news/wonk/wp/2017/01/25/here-...
Expanding on your point, the biggest risk I see with postcards rather than really plain security envelopes would be that savvy postal workers may become acquainted with what even ambiguous or anonymous-sender postcards would look like, enabling them to literally deny service (acting, again literally, as a web service DoS) to particular street addresses making election ad buys by suppressing delivery. With time-critical ad purchases just weeks or even potentially <7 days before election day, this would seem like a legal pandora's box, and even anonymizing postcards wouldn't sufficiently mitigate the risk of an employee blackballing a postcard with an exposed PIN or scratch-off on it that's heading e.g. to Koch HQ or MoveOn.org within weeks of election day.
It'll have to go to a residential / business address
How about "1050 Kiely Blvd #666, Santa Clara CA 95055"?That's actually a PO Box. (Replace 666 with a real number) Do you think Facebook would care?
They should also require the buyer take a photo saluting the American flag while eating a hamburger.
"“If you run an ad mentioning a candidate, we are going to mail you a postcard and you will have to use that code to prove you are in the United States,” Harbath said at a weekend conference of the National Association of Secretaries of State, where executives from Twitter Inc and Alphabet Inc’s Google also spoke.
“It won’t solve everything,” Harbath said in a brief interview with Reuters following her remarks.
But sending codes through old-fashioned mail was the most effective method the tech company could come up with to prevent Russians and other bad actors from purchasing ads while posing as someone else, Harbath said."
Less postcards, more unraveling shell corporations using their corporate metadata. Authenticate political ad buyers using LexisNexis, IRS tax IDs/political org registrations, perhaps going so far as to implement KYC [1] controls and procedures. It's not going to scale, plan on hiring if this matters. Otherwise, this is going to be defeated with a UPS store mailbox.
"KYC controls typically include the following:
* Collection and analysis of basic identity information such as Identity documents (referred to in US regulations and practice as a "Customer Identification Program" or CIP)
* Name matching against lists of known parties (such as "politically exposed person" or PEP)
* Determination of the customer's risk in terms of propensity to commit money laundering, terrorist finance, or identity theft
* Creation of an expectation of a customer's transactional behavior
* Monitoring of a customer's transactions against expected behavior and recorded profile as well as that of the customer's peers"
[1] https://en.wikipedia.org/wiki/Know_your_customer
Source: I work in risk management for a financial services firm.