New Delhi: Instagram users were taken aback recently as they came across disturbing, sensitive, or violent content.
Disturbing images and explicit content flooded several users’ Reels feed. Among those were violent sexual attacks, grievous injuries, and many other inappropriate videos which baffled users.
Not just India, millions of users across the world got such explicit not safe for work (NSFW) content on Instagram.
After Google search on Thursday revealed that hundreds of Instagram users around the world reported a sudden surge of violent and explicit content in their Reels recommendations on February 26, Meta apologised, and said the “error” has been fixed.
Interestingly, a similar issue had happened exactly two years ago, on February 26, 2023.
Then also, users had reported finding violent videos, including shootings and torture, in their feeds—without having followed or clicked such content earlier.
Meta had admitted the issue in 2023, claiming that such videos were a small fraction of total views.
“Two years ago, the same exact violent content in gore-filled feed was a big enough deal that the Washington Post wrote an article on it. Only question now is what causes the algorithm to promote this content on February 26 in particular?” asked a Reddit user.
Meta apologised for the latest problem, without addressing the coincidence of such an issue on the same date two years ago.
“We have fixed an error that caused some users to see content in their Instagram Reels feed that should not have been recommended. We apologise for the mistake,” a Meta spokesperson told CNBC.
Mark Zuckerberg’s Meta has always said its policies designed to protect users from graphic imagery.
Meta’s moderation system relies on artificial intelligence, machine learning, and human reviewers to detect and remove harmful content.
However, the latest incident has raised questions about whether these systems are working appropriately.
This issue comes weeks after Meta announced a significant shift in its content moderation approach.
The company said on January 7 that it would henceforth focus more on serious violations, such as terrorism, child exploitation, fraud, and scams, while relying on user reports for less severe violations.
