Meta Admits Wrongly Suspending Facebook Groups But Denies Wider Problem

Meta says it is "fixing" a problem which has led to Facebook Groups being wrongly suspended - but denied location is simply a wider problem pinch its platforms.

In online forums, Group administrators opportunity they person received automated messages stating, incorrectly, that they had violated policies truthful their Groups had been deleted.

Some Instagram users person complained of akin problems pinch their ain accounts, pinch galore blaming Meta's artificial intelligence (AI) systems.

Meta has acknowledged a "technical error" pinch Facebook Groups, but says it has not seen grounds of a important summation successful incorrect enforcement of its rules connected its platforms much widely.

One Facebook group, wherever users stock memes astir bugs, was told it did not travel standards connected "dangerous organizations aliases individuals," according to a station by its founder.

The group, which has much than 680,000 members, was removed but has now been restored.

Another admin, who runs a group connected AI which has 3.5 cardinal members, posted connected Reddit to opportunity his group and his ain relationship had been suspended for a fewer hours, pinch Meta telling him later: "Our exertion made a correction suspending your group."

It comes arsenic Meta faces questions from thousands of group complete nan wide banning aliases suspension of accounts connected Facebook and Instagram.

A petition entitled "Meta wrongfully disabling accounts pinch nary quality customer support" has gathered almost 22,000 signatures astatine nan clip of penning connected change.org.

Meanwhile, a Reddit thread dedicated to nan rumor features galore group sharing their stories of being banned successful caller months.

Some person posted astir losing entree to pages pinch important sentimental value, while others item they had mislaid accounts linked to their businesses.

There are moreover claims that users person been banned aft being accused by Meta of breaching its policies connected kid intersexual exploitation.

Users person blamed Meta's AI moderation tools, adding it is almost intolerable to speak to a personification astir their accounts aft they person been suspended aliases banned.

BBC News has not independently verified those claims.

In a statement, Meta said: "We return action connected accounts that break our policies, and group tin entreaty if they deliberation we've made a mistake."

It said it utilized a operation of group and exertion to find and region accounts that collapsed its rules, and was not alert of a spike successful erroneous relationship suspension.

Instagram states connected its website AI is "central to our contented reappraisal process". It says AI tin observe and region contented against its organization standards earlier anyone reports it, while contented is sent to quality reviewers connected definite occasions.

Meta adds accounts whitethorn beryllium abnormal aft 1 terrible violation, specified arsenic posting kid intersexual exploitation content.

"We return action connected accounts that break our policies, and group tin entreaty if they deliberation we've made a mistake," a spokesperson added.

The societal media elephantine besides told nan BBC it uses a operation of exertion and group to find and region accounts that break its rules, and shares information astir what action it takes successful its Community Standards Enforcement Report.

In its past version, covering January to March this year, Meta said it took action connected 4.6m instances of kid intersexual exploitation - nan lowest since nan early months of 2021. The adjacent version of nan transparency study is owed to beryllium published successful a fewer months.

Meta says its kid intersexual exploitation policy relates to children and "non-real depictions pinch a quality likeness", specified arsenic art, contented generated by AI aliases fictional characters.

Meta besides told nan BBC it uses exertion to place perchance suspicious behaviours, specified arsenic big accounts being reported by teen accounts, aliases adults many times searching for "harmful" terms.

This could consequence successful those accounts not being capable to interaction young group successful future, aliases having their accounts removed completely.