'Zionist' Google search responses.
(photo credit: screenshot)
Advertisers have been able to target Facebook users who expressed hatred for Jews in their profiles and Google searchers who typed in antisemitic keywords, both companies acknowledged on Sunday in response to media probes.
Google’s algorithm offered antisemitic keywords that advertisers could use to target people making antisemitic and racist searches, a Google employee told The Jerusalem Post.
Some of the predicted keywords included questions such as “The evil Jew,” “Jewish control of banks” and “Why do Jews ruin everything?” The company’s advertisement-buying platform, Google AdWords, then tracked how many viewers saw the antisemitic ads.
The search engine, which handles more than two-thirds of all online queries in the US, said it disabled a majority of the offensive keywords and pulled all the advertisements.
But a quick search showed that some of the predicted keywords
were still on the site.
“We cannot make sure 100% that this won’t happen again. The algorithm changes, the way people search for stuff changes. Nothing is bulletproof but we’re working on it,” the Google employee said.
On average, Google reviews tagged items within one business day, said the employee, adding that some 1.5 billion ads have been removed by the company for violating its policies. The company is now maintaining a stricter blacklist of words that require human approval – as opposed to an automated advertising algorithm’s approval – in order to avoid a repeat of past problems.
The world’s dominant search engine and its largest social media platform – Google and Facebook – dominate online advertising by offering a trove of personal data that advertisers can use to target users.
Google’s controversy came one day after ProPublica reported that Facebook generated advertisement categories such as “Jew hater,” “How to burn Jews,” and “History of why Jews ruin the world.” The story was investigated through ProPublica’s purchase of antisemitic ads on the social media portal.
The number of Facebook users who would have been targeted by those Facebook advertisements was too small to process an ad buy, so ProPublica specified targets such as the SS, the Nazi Party, and Germany’s far-right National Democratic Party. The news site paid $30 for three “promoted posts” and Facebook approved the advertisement requests within 15 minutes. The ads eventually reached some 5,900 people.
Facebook also automatically proposed that “Second Amendment” supporters be shown the ads, possibly because its algorithm found a correlation between gun hobbyists and antisemites.
In response, the social-media giant said it would temporarily bar advertisers from targeting users based on people’s education and employer fields. “We are removing these self-reported targeting fields until we have the right processes in place to help prevent this issue,” the company said in a statement.
Facebook then removed the flagged antisemitic categories and said it would start to screen and remove offensive content more quickly.
“Hate speech and discriminatory advertising have no place on our platform,” Facebook said in a statement.
“Advertisers can report any inappropriate targeting fields directly in the ads interface or via our Help Center.”
It is unclear how the two companies decide what constitutes offensive content in need of removal, raising questions of free speech. When asked, representatives from both Google and Facebook demurred, saying it was an internal matter and that both firms maintained detailed hate-speech policies.
To help prevent a repeat of the antisemitic advertisements, Facebook said it would expand its monitoring and screening of posts that are flagged as inappropriate by other users.
“Over the next year, the company will be adding 3,000 people to the community operations team around the world – on top of the 4,500 Facebook has today – to review the millions of reports we get every week and add more reporting tools for the users,” Facebook spokeswoman Julie Jaffe wrote in an email to the Post