The anniversary of its home country’s birth was not without incident for Facebook, which attracted attention for automatically striking down a post, flagged as hate speech, containing the words of the Declaration of Independence.
Commentators conjecture that the historic document’s reference to Native Americans as “merciless Indian Savages” may have had something to do with the alarm system’s response.
Facebook restored the post and owned up to its system’s continued inability to review without error millions of posts each week and correctly identify which ones should be taken down for hateful content, Quartz reported.
But the incident is of apiece with what has become a redundant reminder: Facebook’s post review system, which combines manpower and AI-backed evaluation, seemingly cannot do its job at such a fantastic scale without generating press-worthy errors.
The social giant played down the significance of that fact, noting that it swiftly addressed the report of an erroneous post review and that the media organization that published the post uses its platform for free.
Yet for media organizations anxious about such errors, last week’s story is more fodder for the argument that smaller, local organizations are prime potential content and advertising partners for media creators and buyers fretful about Facebook’s trustworthiness.