How algorithms are fuelling anti-Muslim hate globally

Most of the anti-Muslim hate generated online is amplified through fake or algorithm-generated accounts, a study has revealed. Such tweets were found to often target Muslim public figures

Lawrance Pintak, a former journalist, who headed the research in July 2021, analysed tweets directed toward Muslim US representative Ilhan Omar during her campaign. The research found that at least half of the tweets were ‘overtly Islamophobic and xenophobic and other forms of hate speech’.

Most of the hateful content was put out by “provocateurs”, conservatives who spread anti-Muslim hate. Such conversations weren’t being boosted by the provocateurs themselves. Those were being amplified by users who pushed the posts by provocateurs. It was also aided by accounts using fake identities in a bid to manipulate conversations online. Pintak termed such accounts as “sockpuppets”.

The study further revealed that out of the top 20 anti-Muslim websites only four were authentic accounts that were inflaming anti-Muslim rhetoric. The rest of these were bots generated through algorithms, reported TRT World.

How AI is being used to amplify anti-Muslim hate

In recent times, Generative Pre-trained Transformer 3 (GPT-3) an Artificial Intelligence system, has been used to circulate anti-Muslims.

It includes and perceives stereotypical misconceptions about Islam. Expressing views on the GPT-3 Abubakar Abid, founder of Gradio — a platform for making machine learning accessible said, “I am shocked how hard it is to generate content regarding Muslims from the system which is not related to violence.”

In order to understand if anti-Muslim hate is documented, Abid conducted an experiment. While researching he noticed that this isn’t just a problem with GPT-3. Even GPT-2 suffers from the same bias issues. Abid also found that the AI completed the text by itself. In 2021 Abid along with two other researchers published a paper describing how AI systems such as GPT-3 associate Muslims with violence.

Facebook’s anti-Muslim algorithms

In 2019, Snopes conducted an investigation which revealed, how certain right-wing Christians manipulated Facebook algorithms through anti-Muslim pages. This was done to establish a coordinated pro-Trump network that spread hate and conspiracy theories about Muslims.

These pages claimed Islam was not a religion and depicted Muslims to be violent. They further claimed that the influx of Muslim refugees in Europe is a “cultural destruction and subjugation”. All of this anti-Muslim hate was ignored by the social media giant.