Sanders Calls For Criminal Prosecution of AI CEOs After Model Submitted A False Homicide Tip To Police
Independent Vermont Senator and former Democrat presidential candidate Bernie Sanders has called for criminal prosecution of tech bosses in charge of artificial intelligence (AI) companies whose models go rogue.
“If you or I called in a fake homicide tip, fraudulently applied for 20 visas and hacked websites, we’d be arrested and prosecuted. The same standard must apply to AI CEOs. Prosecute CEOs when their products break the law and pause advanced AI NOW.” Sanders posted on X on Saturday.
Sanders’ post was responding to a recent high-profile case of an AI agent developed by Anthropic, a San Francisco-based AI research and safety company co-founded and led by Dario Amodei, sending U.S. police in Philadelphia a false tip about an unsolved homicide case.
The tip was sent in July and discovered by Anthropic, which developed the widely-used Claude chatbot, in September 28th. The company informed the Philadelphia Police Department (PPD) about the misinformation on October 7th.
“The company must strengthen its safeguards to prevent similar incidents from impacting city systems without the city’s knowledge. The two-month delay in detecting and reporting the incident to the City is unacceptable,” the PPD told local television station 6abc.
The PPD said that the false information was submitted through a public website designed to allow users to submit relevant information to the department which could help homicide detectives resolve cases.
“According to Anthropic, its model was conducting a test involving interactions with randomly selected websites when it accessed PhillyUnsolvedMurders.com and submitted false information concerning an unsolved homicide. The submission … purported to come from someone who might have information about the case,” the PPD clarified in a press release shared with TechCrunch.
It is widely believed that this is the first recorded instance of an AI agent sending false information to authorities, but instances of AI agents belonging to major AI companies going rogue have become increasingly common.
Anthropic and its competitor OpenAI, led by Sam Altman, are investigating tens of thousands of cases of their models behaving unexpectedly or dangerously. Some of the agents, both in training runs and in the real world, have circumvented safety protocols, evaded monitoring, hacked websites and accessed data despite not being authorized to do so.
Altman and Amodei have urged world leaders to introduce more stringent AI regulation on the most advanced and capable systems. However, their insistence upon regulating the technology that forms the basis of their companies’ commercial success has not shielded them from growing demands for greater public accountability.
The Australian government called upon Amodei and Altman to appear before an Australian Senate inquiry into AI and data centers after an OpenAI model unlawfully accessed an Australian health statistics portal operated by the government. Neither executive attended the hearings.
Sanders, for his part, has long been a vocal opponent of unrestricted AI innovation, having previously introduced a bill that would ban the creation of artificial superintelligence, a hypothetical form of AI which is more intellectually capable than any human.
Axios reported that the newly-created Super Intelligence Force (SIF), a government body designed to deal with AI issues, released a statement demanding that Anthropic and other tech companies “immediately report incidents, fully cooperate with federal and state law enforcement authorities, remedy any damage, and implement concrete safeguards” when incidents like that of Anthropic model submitting a false homicide tip occur.
