The U.S. Army is about to board a Chinese ship. Failed the study generated by artificial intelligence

The U.S. armed forces almost carried out an armed boarding of a Chinese cargo ship.

Reason? A bot-generated full false intelligence study that misidentified the cargo being carried as secret atomic components.

How She Revealed CNN stationand described the service Ars Technica, the incidental occurred in the spring of that year, during the war with Iran. An analyst of the U.S. peculiar Operations Command (US SOCOM) utilized an unidentified chatbot to analyse the documentation of the cargo of the Chinese unit. The tool combined publically available data with classified electronic intelligence information, as a consequence "inferring" that the ship carries atomic components. The military began preparations to intercept the unit with aviation support.

According to CNN, the analyst utilized artificial intelligence for the second time to turn the findings received into a standard military report. specified a formatted paper went into authoritative circulation and began to circulate in the American military. The mistake was only detected at the last minute, interrupting the action (the CNN did not get the information which was actually on board the ship). 1 CNN origin summed up the situation with the message that the machine's mistake had “almost caused war”.

Pentagon trusts machines

Although it is not known whether an analyst utilized a commercial or government tool, the incidental perfectly illustrates the hazard of AI hallucinations in critical situations. Nevertheless, the U.S. Department of defence (DoD) powerfully accelerates the implementation of large language models. Last December, GenAI.mil's military platform was based on the Gemini for Government system, and last period the Grok model was added to it. A Pentagon typical in June praised legislature that 1.5 million military personnel had already utilized these tools.

American services besides have an adapted version of Claude from Anthropic. Interestingly, in March Department of defence blacklisted the company after that., as she opposed the usage of her models in autonomous weapons systems (last period the national justice declared this decision unlawful retaliation).

The authoritative ban is not enough. U.S. Army continues to usage “forbidden” artificial intelligence from Anthropic

The Declaration of the State Department of 2023 called for the work and presence of a "man in the decision loop" to be maintained erstwhile utilizing AI systems. However, the case with a Chinese ship shows that even with human oversight, trust in a digital assistant can be very deceptive, and a machine-generated paper can easy identify a dangerous error.

If article The U.S. Army is about to board a Chinese ship. Failed the study generated by artificial intelligence does not look right in your RSS reader, then see it on iMagazine.