X's algorithms amplified false claims and racist violence after Southport attack
Amnesty International reported that X's algorithms amplified false claims about the religious and immigration status of the Southport stabbing suspect, contributing to far-right riots and violence against Muslim and migrant communities in the UK. The platform's design prioritizes engagement without adequate safeguards against harm, allowing misinformation to reach millions of users and fueling anti-Muslim and anti-migrant violence that continues as a human rights risk.
What happened
On 29 July 2024, a stabbing attack in Southport, UK, killed three young girls. False claims circulated on X that the suspect was a Muslim asylum seeker; the suspect was actually a British citizen born in Cardiff, Wales.
According to Amnesty International's analysis of X's code and publicly available software, X's content-ranking algorithms 'systematically prioritize' content that sparks outrage and engagement 'without adequate safeguards' to prevent or mitigate harm. An account called 'Europe Invasion' posted false claims that the suspect was 'alleged to be a Muslim immigrant,' garnering over four million views, with all X posts speculating the perpetrator was Muslim, a refugee, a foreign national, or arrived by boat tracked to have an estimated 27 million impressions within 24 hours.
The algorithmic design appeared to have no mechanism for assessing potential harm from posts as long as they drove engagement, 'at least not until enough users themselves report it,' according to the research.
Reported impact
Far-right riots broke out across the UK following the attack, fueled by the false online claims spread on X. Many mosques, Islamic buildings, and hotels housing migrants were targeted across the country.
Amnesty International stated that X's design and policy choices 'contributed to heightened risks amid a wave of anti-Muslim and anti-migrant violence observed in several locations across the UK last year, and which continues to present a serious human rights risk today.'
Index analysis
X's recommender system prioritizes content based on engagement and emotional reactions such as anger and heated debate without effective mechanisms to mitigate harm, creating what Amnesty International described as 'fertile ground for inflammatory racist narratives to thrive.'
The platform's policy and design choices had no adequate safeguards to prevent the amplification of false narratives about the suspect's identity and immigration status, allowing misinformation to reach millions of users rapidly.
Governance controls
Pat de Brún, Head of Big Tech Accountability at Amnesty International, stated that 'the current design of X poses serious human rights risks' and warned that the platform's emphasis on engagement can jeopardize user safety and rights.
Amnesty International stressed that X's existing policies and platform design could exacerbate violence and hatred toward religious and immigrant minorities in similar crises.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.