OpenAI Publishes National Security Principles for AI Military and Intelligence Use
A small burst of praise on X focused on the limits OpenAI says it will keep around government and military uses of its models.
On X, researcher Andreas Kirsch pointed to OpenAI's government and national security partnerships page and praised what he described as the company's newly published national security principles. A linked X post from Andrew Curran summarized the restrictions as barring uses including mass domestic surveillance and high-stakes decisions or force without human judgment and accountability. Kirsch's post also contrasted OpenAI's language with Google's earlier AI-principles changes.
Kudos to OpenAI actively trying to do the right thing and not be evil. Imagine if Google still were that way.
Commentary on X
Highest ranked@AndrewCurran_ @Miles_Brundage Also love that OAI had hired the specific advisor that the government had to abuse and work around those same laws post snowden. And ex nsa directors on their board. They know exactly what they are doing and what to say to pretend they are doing “good
Combined views
45.5K
5 posts, first seen 17h ago