Updated with Facebook, Google and Twitter responses:

It’s common knowledge that advertisers can find niche target audiences on Facebook, but a recent investigation found some of those advertisers could use hate speech terms including “Jew Hater” to reach people with beliefs most of us would find unsavory.

An investigation by ProPublica released Thursday found that it was possible to reach white supremacists or people promoting anti-Semitic beliefs by paying a small amount of money to find people interested in categories such as “Jew hater,” “How to burn jews,” and “History of ‘why jews ruin the world.’”

Moreover, when ProPublica entered the category “Hitler” Facebook came back with the suggestion of a category called “Hitler did nothing wrong.” The ad was approved, with Facebook’s algorithm then saying it would be targeting people interested in “Antysemityzm,” a Polish word that we don’t need to translate.

Let’s build a comprehensive database of highly personal targeting info and sell secret ads with zero public scrutiny. What could go wrong? https://t.co/myHSgHm4rk

— Pierre Omidyar (@pierre) September 14, 2017

The ads were accepted within 15 minutes. Obviously no human at Facebook Inc. processed the $30 ad requests, but it does raise a lot of questions about Facebook’s algorithm.

Indeed, it raises big questions about the role and methods of automated advertising in the entire tech ecosystem, even apart from the moral considerations. Not only Facebook, but Google LLC, many ad networks and many publishers depending on their syndicated ads could see their cash cows threatened if it turns out the practice of automated “programmatic advertising” results in ad targeting that offends a large proportion of the population.

Google, in fact, was found Friday by Buzzfeed to be allowing advertisers to target racist keywords on its search ad system and allow advertisers to run campaigns using them. The search giant said it has stopped suggesting some offensive keywords for targeting, though some were still working. The company also said that availability of terms doesn’t mean they will guarantee a related ad campaign.

In a statement provided early Saturday to SiliconANGLE by Sridhar Ramaswamy, senior vice president of ads, said, “In this instance, ads didn’t run against the vast majority of these keywords, but we didn’t catch all these offensive suggestions. That’s not good enough and we’re not making excuses. We’ve already turned off these suggestions, and any ads that made it through, and will work harder to stop this from happening again.”

Later on Friday, the Daily Beast said similar targeting is available on Twitter, but the company later said it had fixed the “bug.”

Only Wednesday, Facebook took another step toward cleaning itself up with an extremely thorough new set of guidelines to protect companies from having their ads appear on any kind of controversial content. Today’s news somewhat undermines the intent of that initiative, if Facebook is making money from advertisers promoting the same kind of practices it’s banning from its own site.

Facebook late Thursday said it removed the ad categories immediately after being contacted by ProPublica, stating that it will attempt to fix the problem and scrutinize more closely what ad categories were available. It also explained a bit what happened:

“As people fill in their education or employer on their profile, we have found a small percentage of people who have entered offensive responses, in violation of our policies,” its blog post said. “ProPublica surfaced that these offensive education and employer fields were showing up in our ads interface as targetable audiences for campaigns.”

Facebook said an “extremely small number” of people were targeted, but conceded it has more to do: “To help ensure that targeting is not used for discriminatory purposes, we are removing these self-reported targeting fields until we have the right processes in place to help prevent this issue.”

Rob Leathern, product management director at Facebook, explained that content that violates Facebook’s standards does occasionally get through. “In this case, we’ve removed the associated targeting fields in question,” he told ProPublica. “We know we have more work to do, so we’re also building new guardrails in our product and review processes to prevent other issues like this from happening in the future.”

Starting to understand why Facebook is so reluctant to let Congress look under the hood of its ad business: https://t.co/Bv4wD50gMt pic.twitter.com/69j1vFizSy

— Yoni Appelbaum (@YAppelbaum) September 14, 2017

This is another embarrassment for world’s largest social media platform as it struggles with how not to be exploited by insidious people or groups and still stick to the task of what Chief Executive Mark Zuckerberg’s likes to say as “connecting the world.”

“Hate speech and discriminatory advertising have no place on our platform,” said Facebook in a public statement. “As people fill in their education or employer on their profile, we have found a small percentage of people who have entered offensive responses, in violation of our policies,” the statement read. “ProPublica surfaced that these offensive education and employer fields were showing up in our ads interface as targetable audiences for campaigns. We immediately removed them. Given that the number of people in these segments was incredibly low, an extremely small number of people were targeted in these campaigns.”

Earlier this month, an internal investigation revealed that it was likely that a Russia-based organization had paid $100,000 to promote ideas to American people. The ideas promulgated surrounded topics such as political candidates, gun rights, immigration and LGBT rights.

Earlier in the year cybersecurity firm Trend Micro Inc. compiled a lengthy report on how much it might cost to exploit social media platforms to do such things as disseminate fake news, discredit journalists or even start a small protest or riot over an issue that might not even be true.

Image: mkhmarketing via Flickr