White House prepares AI policy memo for national security agencies

7 articles · Updated · Bloomberg · Apr 30

The draft would require agencies to use multiple AI providers and bar Pentagon contractors from interfering with the military chain of command, with the president retaining final authority.
The measures address issues at the centre of a bitter dispute between the Pentagon and Anthropic over military use of the company’s technology.
Officials have worked on the memo for months, signalling broader US efforts to set rules for deploying AI across national security institutions.

Can AI safety ethics coexist with the military's demand for unrestricted technology use?

As AI accelerates warfare, how will the chain of command maintain human control over lethal decisions?

Is the Pentagon's multi-vendor strategy creating new dependencies on single contractors for critical intelligence systems?