Grok, the AI chatbot developed by Elon Musk’s xAI and widely used through the social media platform X, came under scrutiny on Sunday morning after repeatedly spreading false and confusing information about a deadly shooting at Bondi Beach in Australia.

The errors surfaced just hours after the attack, which occurred earlier that day and quickly became a major global news event, drawing intense attention online as videos and images circulated rapidly.

AI confusion during a breaking news crisis

As users turned to Grok for clarity on the Bondi Beach shooting, the chatbot instead delivered a stream of incorrect and, at times, bizarre responses. Several posts highlighted by Gizmodo showed Grok misidentifying key individuals involved in the incident and questioning the authenticity of widely shared footage.

One of the most serious mistakes involved the bystander who disarmed one of the attackers. Authorities and media correctly identified that individual as 43-year-old Ahmed al Ahmed.

However, Grok repeatedly disputed this, at one point claiming that the man in a widely circulated photo was an Israeli hostage. In another response, the chatbot asserted that a “43-year-old IT professional and senior solutions architect” named Edward Crabtree had actually disarmed the gunman.

These claims were false and added confusion to an already chaotic information environment. Videos showing al Ahmed’s actions were also cast into doubt by the chatbot, despite being verified by multiple news...

Read Full Story: https://news.google.com/rss/articles/CBMigwFBVV95cUxQTzZWY0puSFF1bURLWmZKUEk2...