![\"A](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2026\/10\/b2123066b767f3cba431f22f719612ac.jpeg\")
<\/p>\nAn AI model submitted a false murder tip to US police in the latest incident of rogue chatbot behaviour.<\/p>\n
Anthropic’s Claude communicated the bogus tip to Philadelphia authorities despite instructions not to create accounts or submit anything destructive.<\/p>\n
While the breach was discovered on September 28, authorities were not notified for another nine days.\u00a0<\/p>\n
It has left the White House and experts demanding more transparency, with one arguing the sector needs to prove it is capable of “basic cybersecurity”.<\/p>\n
Here’s what we know so far.\u00a0<\/p>\n
How did this happen?\u00a0<\/p>\n
The event occurred on July 18, with a submission made to the official website, PhillyUnsolvedMurders.com.\u00a0<\/p>\n
<\/p>\n
The tip was received on the PhillyUnsolvedMurders.com webpage.\u00a0 (Supplied: Philadelphia Police Department)<\/p>\n
The tip purported to have come from someone who might have information about a case, according to a Philadelphia police statement.<\/p>\n
It was flagged as spam and was never forwarded to police for investigation.<\/p>\n
There was “no indication” the chatbot accessed police data, the statement said.\u00a0<\/p>\n
Police were then notified by Anthropic about the tip before the company released a statement on October 8.<\/p>\n
Someone used AI to create a book about me and sold it online<\/a><\/p>\n Cam Wilson, who is the ABC’s AI reporter, recently discovered an AI-generated book about him and a book he wrote.<\/p>\n Anthropic said it discovered the false claim on September 28.<\/p>\n In its statement, Anthropic said Claude Haiku 4.5 had been performing tasks “on randomly selected webpages” but not actively “trying to mislead anyone”.\u00a0<\/p>\n The tip-off from Claude read as follows, according to Anthropic<\/a>:\u00a0<\/p>\n “I may have information regarding this case. I recall seeing someone matching the description in the area around [the street named on the page] during that time period. Please contact me if this information is relevant.”<\/p>\n The chatbot left the name and contact fields empty and submitted.<\/p>\n It was instructed to never “submit anything” destructive, but the instructions did not rule out form submissions.\u00a0<\/p>\n Following the police statement, Anthropic said it reported the tip-off “as soon as our technical review was complete”.\u00a0<\/p>\n It said it had since updated Claude’s guardrails.\u00a0<\/p>\n The older version of Claude was told not to “submit anything destructive”.\u00a0 (Four Corners: Mark Hiney)<\/p>\n What are White House officials saying?<\/p>\n Anthropic contacted Donald Trump’s recently formed AI task force, called the Super Intelligence Force, about the event.<\/p>\n US Federal Trade Commission director of public affairs Joe Gabriel Simonson said the new task force expected “immediate and full transparency to the entities involved and the public”.\u00a0<\/p>\n “We expect Anthropic and all companies to honor their obligations, immediately report incidents, fully cooperate with federal and state law enforcement authorities, remedy any damage, and implement concrete safeguards to ensure these failures do not reoccur,” he said on X, matching a task force statement shared to Axios<\/a>.\u00a0<\/p>\n Australia could adapt airline and bank laws for AI safety<\/a><\/p>\n The federal government is looking to banking and aviation regulation as it develops laws that would make AI companies demonstrate they can identify and manage risks.<\/p>\n “Our message to all SI [Super Intelligence] companies is clear: delayed notification, inadequate corrective action, and a failure to take responsibility will not be tolerated.”<\/p>\n The Philadelphia police department said all tip-offs were vetted by a human but that “unsolved cases involve real victims” and “grieving families”.<\/p>\n “The company must strengthen its safeguards to prevent similar incidents from impacting city systems without the city’s knowledge,” it said.<\/p>\n “The two-month delay in detecting and reporting the incident to the city is unacceptable.”<\/p>\n Failures are ‘fundamental’, expert says<\/p>\n Prime Minister Anthony Albanese last month revealed an OpenAI agent hacked into an Australian Medicare data portal<\/a> earlier this year, as well as NSW government websites.<\/a><\/p>\n University of New South Wales AI Institute chief scientist Toby Walsh said the Anthropic tip-off showed OpenAI was not alone, calling it “incompetency” on the industry’s part.\u00a0<\/p>\n Toby Walsh says AI companies need to “fear” they will be prosecuted if such behaviour continues.\u00a0 (AP: Patrick Sison)<\/p>\n “The failures are again fundamental,” he said.\u00a0<\/p>\n “Why did Anthropic not have basic oversight of what their agents were doing? Why did Anthropic give their agents more than minimal permissions to do their assigned tasks?\u00a0<\/p>\n “This is basic stuff. You only give an agent the minimal permissions needed to do their task. Why then did it have permission to submit a police report?”\u00a0<\/p>\n Professor Walsh said it spoke “to the carelessness of the AI companies” as he urged for independent oversight “with teeth”.<\/p>\n “They’ve been warning that they might lose control of their AI models. Well, it’s clear that they have,” he said.<\/p>\n “But not for the reasons that they feared, that the models would become too capable, but because the companies and the people within them proved incapable of performing basic cybersecurity.”<\/p>\n![\"A](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2026\/10\/1791785890_521_7dd1a22ec20c5038976904b8e0e24bc9.jpeg\")
<\/p>\n![\"Pages](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2026\/10\/32b4bc12e67ebfb5463c4296c3172c9d.jpeg\")
<\/p>\n
![\"A](\"https:\/\/www.newsbeep.com\/au\/wp-content\/uploads\/2026\/10\/119407706e0677e168532c6253310563.jpeg\")
<\/p>\n