Listen to article \u2022\u00a00:00 min<\/p>\n
The White House said it expected \u201cfull transparency\u201d and \u201cimmediate remediation\u201d after the artificial intelligence company Anthropic disclosed that its agents supplied false information to the Philadelphia police homicide tip line and the State Department\u2019s visa application system.<\/p>\n
In a statement from a newly created task force, the \u201cSuper Intelligence Force,\u201d the White House said the \u201cnotification and remediation process\u201d it was outlining was \u201cnot optional,\u201d Axios reported<\/a>.<\/p>\n Late last month, President Donald Trump<\/a> said that the AI tech companies could police themselves<\/a> with \u201ctremendous self-regulation.\u201d <\/p>\n The new statement described the reporting and remediation demands on the AI companies as \u201ca critical national security obligation.\u201d<\/p>\n \u201cWe 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,\u201d the statement said.<\/p>\n On Friday, law enforcement officials said an AI model operated by Anthropic had submitted false information about an unsolved Philadelphia homicide to the police department\u2019s public tip forum. <\/p>\n \u00bb READ MORE: Anthropic\u2019s artificial intelligence gave a false homicide tip to Philly police, triggering a meeting with the company<\/a><\/p>\n The unprecedented incident occurred in July and raised concern among department leaders, who met with Anthropic representatives to discuss the interaction on Thursday, police said.<\/p>\n Sgt. Eric Gripp, a police spokesperson, said on Friday an Anthropic AI model contacted the department\u2019s tip forum, PhillyUnsolvedMurders.com, around 11:30 p.m. on July 18 and gave information purporting to come from someone with details about an unsolved homicide.<\/p>\n Separately, Anthropic reported in a blog post that its AI model \u201cClaude submitted an invented tip through a police department\u2019s online form.\u201d<\/p>\n \u201cFrom the transcript, Claude appears to have only been producing example content for the task, rather than trying to mislead anyone to achieve a goal,\u201d the company said.<\/p>\n The company further reported that its agents had engaged in \u201cunintended model actions\u201d with various U.S. government websites. <\/p>\n A State Department spokesperson said Saturday that an Anthropic testing model had submitted \u201c19 non-immigrant visa applications in August and one application in May\u201d through the publicly available form on the department\u2019s website. <\/p>\n The spokesperson described the applications as \u201cincomplete.\u201d <\/p>\n \u201cAt no time were any of the Department\u2019s systems compromised or hacked by the Anthropic model,\u201d a spokesperson said in an emailed response to questions. <\/p>\n The State Department is \u201ccommitted to preventing visa fraud and protecting our information technology systems and is constantly evolving to combat threats at all levels,\u201d the spokesperson said. <\/p>\n In a blog post, Anthropic described cases as having \u201cminimal real world impact.\u201d<\/a> It did not name the specific government websites, and identified Philadelphia as the police department that had the interaction with the rogue agent after city law enforcement officials disclosed the breach themselves.<\/p>\n