Is Meta’s AI “mistakenly” censoring too much? Nick Clegg, the company’s top manager, says yes

Meta’s top manager sheds light on the company’s moderation efforts. This also includes the AI ​​incorrectly removing too much content across all platforms.

Nick Clegg, head of global operations at the tech giant, admitted that many mistakes continue to be made when it comes to removing content he deems “unnecessary”. He also acknowledged that error rates remained too high, but called for changes to improve accuracy and precision.

Clegg also talked about how the company may try to follow all the rules in the book and as a result, innocuous content will be removed. In the long run, this leads to many users being unfairly penalized.

The company has come under scrutiny in the past for deleting numerous materials related to the COVID-19 pandemic. Zuckerberg admitted that many bad decisions were made after he was under pressure from the government.

Back then, the rules were stricter and the timing wasn’t optimal for the pandemic. So it came down to wisdom and poor decision making. They admitted that they overdid it quite a bit. And it’s thanks to Meta app users who raised their voices to be heard.

These comments are evidence that Meta’s automated AI systems are simply unnecessarily harsh. It’s a good example of moderation errors that were once commonplace in threads. But even though the company knows everything, it has not yet made any major substantive changes since the election began.

On the other hand, Meta’s oversight board, ostensibly set up to address complex moderation dilemmas, remains powerless to address all shadow bans and over-moderated content due to the lack of a reporting mechanism and restrictions. Critics argue that this is not a mere oversight, but a calculated move to limit the board’s influence and keep ultimate control firmly in Meta’s hands. This is consistent with Meta’s controversial track record of suppressing certain voices and amplifying harmful narratives. For example, during the Rohingya crisis, Meta was accused of enabling violence against minorities, and in the Israeli-Palestinian conflict, Meta censored pro-Palestinian and Gaza voices fighting genocide. Such actions suggest that Meta’s commitment to fairness may be secondary to its bottom line, prioritizing profit and power dynamics over accountability and justice.

Does this mean big changes are coming? Well, we think so, because Meta is definitely working on it, especially since the issue is confirmed. Clegg admitted that not many details can be provided as the discussions are at a high level. However, they hope to work with the Trump administration to bring about new changes.

So for users like us, it’s time for Meta to make some necessary moderation changes and give users’ content the recognition it deserves. What do you think?

Image: DIW-Aigen

Read more: Google’s AI-Powered Store Reviews: A Game-Changer for Shoppers or a Nightmare for Businesses?