After removing self-reported targeting fields last week, Facebook today announced a broader set of long-term measures to improve ad targeting on the platform.
The move was prompted by an investigation by the news outlet ProPublica, which revealed that Facebook's algorithm enabled advertisers to serve ads to almost 2,300 "Jew haters."
"Seeing those words made me disgusted and disappointed -- disgusted by these sentiments and disappointed that our systems allowed this," wrote Sandberg. "Hate has no place on Facebook -- and as a Jew, as a mother, and as a human being, I know the damage that can come from hate."
She took responsibility for the blunder, saying that the fact that the categories were available as options for ad targeting was "totally inappropriate and a fail on our part."
"We never intended or anticipated this functionality being used this way -- and that is on us," she wrote. "And we did not find it ourselves -- and that is also on us."
After a manual review of existing targeting options, Facebook has reinstated about 5,000 of the most commonly used targeting terms, said Sandberg, for example "nurse," "teacher" or "dentistry." She said the company would continue to manually review new ad targeting options moving forward to prevent offensive terms from appearing.
Sandberg reaffirmed that content that goes against Facebook's community standards, including anything that directly attacks people based on their race, ethnicity, national origin, religious affiliation or sexual orientation among others, cannot be used to target ads. She added that Facebook was also in the midst of developing a program that people could use to report potential abuses directly.
"We hope these changes will prevent abuses like this going forward," she concluded. "If we discover unintended consequences in the future, we will be unrelenting in identifying and fixing them as quickly as possible."
This post originally appeared on Business Insider.